RESSALVA Atendendo solicitação do(a) autor(a), o texto completo desta dissertação será disponibilizado somente a partir de 23/01/2027. UNIVERSIDADE ESTADUAL PAULISTA - UNESP FACULDADE DE CIÊNCIAS E TECNOLOGIA CAMPUS DE PRESIDENTE PRUDENTE PÓS-GRADUAÇÃO EM CIÊNCIAS CARTOGRÁFICAS LETÍCIA FERRARI CASTANHEIRO A BACKPACK MOBILE LASER SCANNING SYSTEM: FROM SYSTEM DESIGN TO TRAJECTORY ESTIMATION AND POINT CLOUD GENERATION Presidente Prudente/SP 2025 UNIVERSIDADE ESTADUAL PAULISTA - UNESP FACULDADE DE CIÊNCIAS E TECNOLOGIA CAMPUS DE PRESIDENTE PRUDENTE PÓS-GRADUAÇÃO EM CIÊNCIAS CARTOGRÁFICAS LETÍCIA FERRARI CASTANHEIRO A BACKPACK MOBILE LASER SCANNING SYSTEM: FROM SYSTEM DESIGN TO TRAJECTORY ESTIMATION AND POINT CLOUD GENERATION A thesis submitted to the Post-Graduate Program in Cartographic Sciences (PPGCC) at São Paulo State University (UNESP), School of Technology and Sciences, campus Presidente Prudente – SP, Brazil, for the partial fulfilment of the requirements for the degree of Doctor in Cartographic Sciences. Supervisor: Prof. Dr. Antonio M. G. Tommaselli Co-Supervisors: Prof. Dr. Antero Kukko Dr. Mariana B. Campos Presidente Prudente/SP 2025 IMPACTO POTENCIAL DESTA PESQUISA Nesta pesquisa foi desenvolvido um sistema de varredura a LASER móvel e uma técnica para estimar a trajetória e gerar um modelo 3D do ambiente. Este sistema tem potencial para conferir maior eficiência que os métodos convencionais em várias aplicações, como extração de parâmetros de interesse da agricultura digital, entre eles arquitetura do dossel e biomassa. POTENTIAL IMPACT OF THIS RESEARCH In this research, a mobile laser scanning system and a technique for estimating trajectory and generating 3D environmental models were developed. This system has the potential to provide more efficiency than conventional methods in various applications, such as extracting parameters of interest in digital agriculture, including canopy architecture and biomass. To my parents and brother for their love and endless support. To the friends and family, we find along the way. To the experiences we never expect. ACKNOWLEDGMENTS I would like to thank the School of Technology and Sciences, São Paulo State University (FCT- UNESP) and the Post-Graduate Program in Cartographic Sciences (Programa de Pós- Graduação em Ciências Cartográficas – PPGCC) for providing the resources and infrastructure for conducting this research. This study was financed by: • The National Council for Scientific and Technological Development, CNPq (Process number 141550/2020-1) for funding my PhD studies. • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) - Finance Code 001, in the scope of the Program CAPES-PrInt (“From Precision Agriculture to Prescription Agriculture: insertion of UNESP in the context of Digital Agriculture and Bioeconomy”), Process number 88887.310463/2018-00, mobility number 88887.695922/2022-00, which supported my mobility visit to Finnish Geospatial Research Institute (FGI) in National Land Survey (NLS), Espoo, Finland. • The Academy of Finland (337656) by grants “Understanding Wood Density Variation Within and Between Trees Using Multispectral Point Cloud Technologies and X-ray Microdensitometry” (Grant n. 331708) and “Digital Technologies, Risk Management Solutions and Tools for Mitigating Forest Disturbances” (Grant n. 353264), which founded field measurements and participation in international conferences during mobility visit at FGI. • São Paulo Research Foundation, FAPESP, Brazil, Process number 2021/06029-7, for founding the Thematic Project “High resolution remote sensing for digital agriculture”, which supported the field measurements. • The ISPRS Foundation (TIF) for supporting my participation in the ISPRS Congress (Nice, France) and ISPRS TC III Symposium (Belém, Brazil) through travel grants. I am grateful to my supervisor, Prof. Dr. Antonio M. G. Tommaselli, for his guidance and support throughout my academic journey. His experience has been crucial for developing this thesis and my growth as a researcher. I extend my thanks to my co-supervisors: Prof. Dr. Antero Kukko, who provided support and guidance during my mobility visit to FGI; and Dr. Mariana B. Campos, who offered valuable support during my stay in Finland and whose expertise has significantly contributed to this work. Their patience and background have been important to the success of this thesis. Thanks to Nuvem UAV and T2R colleagues for supporting the development of the BLS. I am also grateful to fellow researchers in the Photogrammetry and Remote Sensing research group of FCT-UNESP. Their support has been valuable in field measurements and has provided valuable technical support for this research. I would like to extend my gratitude to all my friends at PPGCC. Our discussion and exchanging experiences enabled us to create an environment for both professional and personal growth. I would like to extend my gratitude to all FGI researchers who contributed to this work by sharing codes and exchanging knowledge, which was crucial for the development of this research. I am also grateful for their warm welcome and support during my mobility visit at FGI in Espoo, Finland. This international experience has not only been important for my research but also provided me with a network of colleagues who have become dear friends. Lastly, I wish to thank my family for their endless support throughout my academic journey, and my dear close friends for their understanding during challenging moments and for celebrating every small victory. Thank you! “Great things are done by a series of small things brought together” -Vincent Van Gogh RESUMO Esta tese apresenta o desenvolvimento e avaliação de um Sistema de Varredura a Laser Móvel (SVLM), acoplado a uma mochila, para o mapeamento de culturas perenes, e em particular, de pomares de laranja. Estes ambientes apresentam desafios particulares na determinação da trajetória da plataforma, devido a obstrução parcial de sinais GNSS e à escassez de feições distinguíveis para o registro entre as nuvens de pontos. O SVLM proposto é composto por um sistema de varredura a LASER Ouster OS0-128, que possui um campo de visada de 360° x 90° e uma IMU integrada. A avaliação inicial do sistema concentrou-se no desempenho do Ouster OS0-128 e sua unidade de medida inercial (IMU), com processamento pelo WebSLAM, solução desenvolvida pela Ouster. Com base nos problemas observados, foram pesquisados e analisados os métodos existentes de registro de nuvens de pontos, resultando na implementação e avaliação de uma abordagem para estimativa de trajetória e mapeamento de pomares da laranja. A metodologia desenvolvida utiliza feições planas para o registro sequencial da nuvem de pontos com estimação dos parâmetros da transformação de corpo rígido pelo método dos mínimos quadrados (MMQ). Os resultados com dados reais mostraram a viabilidade do uso desse SVLM para gerar nuvens de pontos densas em pomares de laranjas, com a estimativa simultânea da trajetória percorrida pela plataforma. Além disso, as limitações e os desafios do uso de um SVLM para o mapeamento de culturas perenes são discutidos sendo sugeridos desenvolvimentos futuros. Palavras-chave: LiDAR; Sistema de Varredura a LASER; extração de feições 3D; registro de nuvem de pontos; mapeamento móvel terrestre; Agricultura Digital. ABSTRACT This thesis presents the development and evaluation of a backpack-mounted mobile laser scanning system (BLS) designed for mapping agricultural environments, focusing on orange orchards. Mapping crops (e.g., orange orchards) with BLS presents challenges in estimating the trajectory due to obstruction of GNSS signals and the few distinct geometric features available for point cloud registration. The proposed BLS integrates an Ouster OS0-128 laser scanner, which has a 360° x 90° field-of-view (FoV) and an internal inertial measurement unit (IMU). The initial evaluation of the system focused on the performance of the Ouster OS0-128 and its inertial IMU. Then, existing point cloud registration methods were evaluated resulting in the implementation and evaluation of an approach for trajectory estimation and point cloud generation. The proposed approach uses planar features for the sequential point cloud registration with the estimation of the trajectory by an incremental least squares method (LSM). The results showed the feasibility of using BLS to generate point clouds in orange orchards with the estimation of the trajectory. In addition, the limitations and challenges of employing a BLS for orange orchard mapping were discussed, providing future recommendations. Keywords: LiDAR; laser scanning system; 3D feature extraction; point cloud registration; terrestrial mobile mapping system; Agricultural Mapping. LIST OF FIGURES Figure 2.1 - Example of wide FoV laser scanning system composed of a solid-state sensor. . 24 Figure 2.2 - (a) Simulation of a mobile acquisition using an Ouster laser scanner system with 1024 rotational steps and a horizontal FoV of 360° and the estimated sensor position as (b) discrete points considering 360° laser frames, (c) discrete points with the correction of the 360° laser frames, and (d) as continuous trajectory. ................................................................ 28 Figure 2.3 - Ouster multi-beam technology: (a) the optical module, (b) the internal architecture and (c) the set of micro-optics. ............................................................................. 30 Figure 2.4 – Beam space options for Ouster sensors: (a) uniform (standard), (b) gradient, and (c) below horizon. ..................................................................................................................... 31 Figure 2.5 – Example of data collected using an Ouster OS0-128 sensor. .............................. 32 Figure 2.6 – The OS0-128 sensor with the 128 beams that are simultaneously emitted in an instant. ...................................................................................................................................... 33 Figure 2.7 – OS0-128 coordinate systems: laser (blue) and sensor (green) in (a) the top view, showing the rotation of the encoder and (b) the lateral view. .................................................. 35 Figure 2.8 - Ouster beam coordinate system in (a) the top view and (b) the lateral view ........ 36 Figure 2.9 - Data acquisition: (a) the BLS system, (b) Ouster OS0-128 laser scanner mounted with 45° inclination, (c) an illustration of the integration of the Ouster sensor with a compute for data storage, and (d) an individual scan obtained with the Ouster OS0-128 laser scanner. .................................................................................................................................................. 38 Figure 2.10 - Location of the test area and the acquisition paths: the entire trajectory is represented by both red and yellow lines (Dataset I), the forward path is the yellow line (Dataset II), and the backward path is the red line (Dataset III). ............................................. 39 Figure 2.11 - Top view of the dense point cloud generated for the test area, indicating the Ground Control Points (GCPs) in yellow and the checkpoints in red, which were utilized for georeferencing in datasets II and III. ........................................................................................ 40 Figure 2.12 - Point clouds, in a local reference system, generated in Web SLAM with datasets (a) I, (b) II and (c) III. The scalar fields represent the coordinate h. In (a) and (b), the scalar field is different from (c) because the beginning of the trajectories for datasets I and II was the same, while dataset III was in the opposite direction. ...................................... 41 Figure 2.13 - Details of objects (car and light post) in the point cloud obtained with datasets (a) I, (b) II and (c) III. ............................................................................................................... 42 Figure 3.1 – (a) ICM-20948 TDK InvenSense IMU, (b) the orientation of axes of rotations, and (c) the orientation of axes for the magnetometer ............................................................... 49 Figure 3.2 - Six stationary positions of the OS0-128 Ouster laser scanner with the internal IMU for calibration measurements. .......................................................................................... 50 Figure 3.3 - Computed translational velocities in (a) X-, (b) Y-, and (c) Z-axis for DCM (Hyyti; Visala, 2015), in dash-dotted red lines, MAD (Madgwick; Harrison; Vaidyanathan, 2011), in green lines, and MAH (Mahony et al., 2009), in dashed blue lines, techniques. ...... 52 Figure 3.4 - Computed translational velocities in the X-, in dash-dotted magenta lines, Y-, in yellow line, and Z-axis, in dashed cyan lines, for the DCM method. .................................. 53 Figure 3.5 - Positional errors along the (a) X, (b) Y, and (c) Z axes obtained with DCM, in dash-dotted red lines, MAD, in green line, and MAH, in dashed blue lines, techniques. ........ 54 Figure 4.1 – (a) the BLS with the Ouster OS0-128 laser scanner, Ricoh Theta S dual-fisheye system and a GNSS antenna, (b) the sensors tilted in 45°, (c) the computer, battery, GNSS receiver and WiFi router placed inside the backpack, and (d) an operator carrying the backpack in an orange orchard. ................................................................................................ 62 Figure 4.2 - The location of the test area and the trajectory of the acquisition, in red line. ..... 63 Figure 4.3 – (a) an orange orchard in a terrestrial view and (a) an orange tree showing the density of the canopy. ............................................................................................................... 64 Figure 4.4 – Laser frame point cloud (a) of an orange orchard in perspective view and (b) an orange tree. ............................................................................................................................... 64 Figure 4.5 – Process flow of the point cloud registrations. ...................................................... 65 Figure 4.6 - Keypoints (in red) detected in the first scan with (a) Uniform Sampling, (b) 3D Harris, and (c) 3D ISS approaches. .......................................................................................... 68 Figure 4.7 - Feature-based registration combining SHOT with (a) US and (b) 3D ISS keypoints. .................................................................................................................................. 70 Figure 5.1 – Workflow for trajectory estimation and point cloud generation using BLS data, including data input (IMU measurement and 3D laser frames), processing (IMU trajectory estimation, keypoints extraction and matching, registration) and final outputs (3D laser point cloud and trajectory). ................................................................................................................ 78 Figure 5.2 – Incremental adjustment to estimation the BLS trajectory. ................................... 85 Figure 5.3 – Keypoints used in the registration of the first pairs. ............................................ 88 Figure 5.4 - The estimated a posteriori standard error of the unit weight (𝜎0) along the estimated trajectory for XY components. ................................................................................. 89 Figure 5.5 – BLS 3D trajectory with coordinate frame orientations during initial 5-second acquisition. ................................................................................................................................ 90 Figure 5.6 – XY trajectory estimated with the proposed approach (red line), WebSLAM (dash-dotted blue line) and GNSS data (dashed green line). .................................................... 91 Figure 5.7 - The trajectory segmented into tree ranges: (a) 0-50 m, (b) 50-150 m, (c) 150- 210 m. ....................................................................................................................................... 92 Figure 5.8 – Discrepancies along the trajectory in (a) X, (b) Y and (c) Z direction. ............... 93 Figure 5.9 – The generated laser point cloud of an orange orchard: top view and two cross- sections (I and II) ...................................................................................................................... 94 Figure 5.10 – Orange tree height measured in the BLS point cloud. ....................................... 94 LIST OF TABLES Table 2.1 – Technical information of Ouster sensors: OSDome, OS0, OS1 and OS2............. 31 Table 2.2 – Detailed information of Ouster OS0 sensor with 128 channels (OS0-128). ......... 34 Table 3.1 - Computed translational position for DCM, MAD and MAH’s solutions and point cloud registration (PCR) after 5 seconds. ................................................................................. 53 Table 4.1 - Total of keypoints detected with UniformSampling (US), 3D Harris, and 3D ISS method. ..................................................................................................................................... 67 Table 4.2 – Mean and standard deviation (std.) of the discrepancies of the transformation parameters estimated with US, 3D Harris, and 3D ISS keypoints. .......................................... 68 Table 4.3 - RMSE of the discrepancies of the transformation parameters estimated with US, 3D Harris, and 3D ISS keypoints. ............................................................................................ 69 Table 4.4 - Total of matches obtained with SHOT and keypoints (US and 3D ISS) and the correct matches based on the 3D distances between points. .................................................... 71 Table 4.5 - RMSE of discrepancies of the transformation parameters estimated with the filtered SHOT matched keypoints. ........................................................................................... 71 Table 4.6 - RMSE of discrepancies of the transformation parameters estimated with ICP and CPD. ......................................................................................................................................... 72 Table 5.1 – Statistics of the incremental registration at each position: average standard deviation of estimated positions from the covariance matrix and the 𝜎0................................. 88 CONTENT CHAPTER 1 – INTRODUCTION ........................................................................................ 16 1.1 OVERVIEW ................................................................................................................... 16 1.2 MOTIVATION ............................................................................................................... 18 1.3 OBJECTIVES ................................................................................................................. 19 1.4 THESIS STRUCTURE AND RELATED PUBLICATIONS ....................................... 19 1.5 RELATED GRANTS AND INTERNATIONAL PARTNERSHIPS ............................ 20 CHAPTER 2 – PLATFORM DESIGN: SYSTEM ASSEMBLY AND FEASIBILITY ASSESSMENT ........................................................................................................................ 21 2.1 OVERVIEW ................................................................................................................... 21 2.2 BACKGROUND ............................................................................................................ 23 2.2.1 Solid-state sensors ................................................................................................... 23 2.2.2 Related works using solid-state laser scanning system ........................................... 24 2.2.3 SLAM and the Ouster WebSLAM .......................................................................... 25 2.3 MATERIALS AND METHODS ................................................................................... 29 2.3.1 Ouster solid-state sensor technical information ...................................................... 29 2.3.2 Ouster OS0-128 sensor ............................................................................................ 33 2.3.3 BLS and data acquisition ......................................................................................... 38 2.3.4 Reference data ......................................................................................................... 39 2.3.5 Data processing ....................................................................................................... 40 2.4 RESULTS AND DISCUSSIONS .................................................................................. 41 2.5 FINAL CONSIDERATIONS ......................................................................................... 43 CHAPTER 3 – ALGORITHMS AND DATA PROCESSING: INERTIAL MEASUREMENT UNIT ....................................................................................................... 44 3.1 OVERVIEW ................................................................................................................... 44 3.2 BACKGROUND ............................................................................................................ 45 3.3 MATERIAL AND METHODS...................................................................................... 48 3.3.1 TDK Invensense ICM-20948 IMU.......................................................................... 48 3.3.2 Accelerometer IMU calibration ............................................................................... 49 3.3.3 Experiments ............................................................................................................. 50 3.4 RESULTS AND DISCUSSIONS .................................................................................. 51 3.5 FINAL CONSIDERATIONS ......................................................................................... 54 CHAPTER 4 – DATA PROCESSING: UPGRADED BLS AND ANALYSIS OF POINT CLOUD REGISTRATION METHODS USING DATA FROM AN ORANGE ORCHARD ............................................................................................................................ 56 4.1 OVERVIEW ................................................................................................................... 56 4.2 POINT CLOUD REGISTRATION METHODS ........................................................... 57 4.3 MATERIALS AND METHODS ................................................................................... 61 4.3.1 BLS setup ................................................................................................................ 61 4.3.2 Data acquisition and test area .................................................................................. 62 4.3.3 Data processing ....................................................................................................... 65 4.4 RESULTS AND DISCUSSIONS .................................................................................. 67 4.4.1 Keypoints assessment .............................................................................................. 67 4.4.2 Assessment of SHOT featured-based matching ...................................................... 70 4.4.3 Direct registration assessment ................................................................................. 71 4.5 FINAL CONSIDERATION ........................................................................................... 72 CHAPTER 5 – DATA PROCESSING: INCREMENTAL TRAJECTORY ESTIMATION AND 3D MAPPING BASED ON PLANAR FEATURES ....................... 74 5.1 OVERVIEW ................................................................................................................... 74 5.2 GEOMETRIC FEATURE-BASED POINT CLOUD REGISTRATION ...................... 76 5.3 PROPOSED PROCESSING WORKFLOW .................................................................. 77 5.3.1 IMU measurements processing ............................................................................... 79 5.3.2 Keypoints and feature extraction using FRIS .......................................................... 79 5.3.3 Mathematical model: Point-to-plane BLS point cloud registration......................... 80 5.3.4 Scan-to-scan adjustment for point cloud registration .............................................. 83 5.3.5 Trajectory estimation by the incremental BLS point cloud registration ................. 85 5.3.6 Point cloud generation ............................................................................................. 86 5.3.7 Experimental Assessment ........................................................................................ 86 5.4 EXPERIMENTAL ASSESSMENTS AND RESULTS ................................................. 87 5.4.1 Dataset ..................................................................................................................... 87 5.4.2 Results: Incremental adjustment assessment ........................................................... 87 5.4.3 Results: BLS trajectory assessment ......................................................................... 91 5.4.4 Results: completeness of point cloud ...................................................................... 93 5.5 DISCUSSION AND CONCLUSIONS .......................................................................... 95 CHAPTER 6 – FINAL REMARKS ...................................................................................... 96 REFERENCES ..................................................................................................................... 99 CURRICULUM VITAE .................................................................................................... 112 16 Castanheiro, L. F. CHAPTER 1 INTRODUCTION 1.1 OVERVIEW Digital agriculture has emerged as a tool to promote sustainable and environmentally friendly agriculture, with an efficient use of land resources, which aligns with the United Nations Sustainable Development Goals (United Nations, 2015). Therefore, many technologies have been developed to support the goals of digital farming and sustainable agriculture. Among the new technologies, laser (light amplification by stimulated emission of radiation) scanning systems, also known as light detection and ranging (LiDAR) systems, have become affordable, producing a three-dimensional (3D) representation of the environments (Wehr; Lohr, 1999; Shan; Toth, 2008; Vosselman; Maas, 2010). Another advantage of this technology is its capability to penetrate open spaces of the vegetation canopy, enabling the mapping of under canopy and fostering the application of laser scanning systems in forest and agricultural environments. Laser scanning systems are typically classified according to the acquisition platform (Rivera et al., 2023): aerial, terrestrial static, and terrestrial mobile. An airborne laser scanner (ALS), can be mounted either in an aircraft or on an unmanned aerial vehicle (UAV), enabling fast mapping over large areas. ALS data is widely used for the generation of digital surface models (DSM), from which digital terrain models (DTM) can be extracted with suitable processing. The terrestrial laser scanner (TLS) uses a stationary platform (e.g., tripod) for scanning, offering higher-resolution point clouds of the scanned objects. A terrestrial mobile platform offers significant advantages in agricultural mapping, combining the mobility of ALS with the terrestrial perspective and higher resolution of TLS. The use of laser scanning systems in agricultural environments started with ALS for monitoring tasks (Davenport; Holden; Gurney, 2004). Then, many algorithms were developed to process ALS data from agriculture fields (Wallace, 2013; Maimaitijiang et al., 2020; Shendryk et al., 2020). However, despite the efficiency of large-scale mapping, the resolution may lack the details level needed for crop analysis. The terrestrial systems have become popular because of their high resolution and ground-level mapping, allowing a different perspective of crops. Therefore, TLS has been applied for the geometric characterisation of trees (Pfeiffer et al., 2018), phenotyping (Jin et al., 2019) and fruit detection (Silva; Santos; 17 Castanheiro, L. F. Galo, 2024). However, applying TLS in larger agricultural areas can be laborious and time- consuming due to the necessity of multiple overlapping scans to generate a complete point cloud of the environment. The terrestrial mobile technology provides faster data acquisition over large areas (Kukko et al., 2012), reducing occlusions. In addition, the ability to navigate using the acquired data, enables its use in autonomous farming equipment, with precise path traversing. The use of mobile laser scanning (MLS) systems has been proposed for canopy density estimation (Lowe et al., 2021), tree height and diameter breast height (DBH) estimation (Zhou et al., 2021; Castanheiro et al., 2024a; Silva et al., 2024), robot navigation (Jiang et al., 2022), and harvesting (Mao et al., 2022). Laser scanning systems also contribute to autonomous farming operations, such as machine auto-driving, seeding, harvesting, plant health, and growth monitoring (Rivera et al., 2023). MLS systems can be mounted on various platforms, including cars, tractors, boats, trolleys, boats, and backpacks (Kukko et al., 2012). Among these, most MLS for agriculture is mounted in tractors (Wang et al., 2021; Zhou et al., 2021; Mao et al., 2022; Pagliai et al., 2022). On the other hand, a backpack-mounted mobile laser scanning system (BLS) offers the advantage of accessing areas inaccessible to wheeled vehicles, such as densely planted orchards or steep terrains. BLS also minimize soil compaction compared to car or tractor platforms. Conversely, the point clouds from a backpack platform suffer from motion disturbances due to the operator movement, and backpack platforms are limited by the operator's speed. The speed to complete the task can be improved by increasing the number of channels, scanning speed, and scanning frequency. Despite BLSs have been widely tested in forest applications (Kukko et al., 2017; Liang et al., 2014; Shao et al., 2020; Zhou et al., 2024), its use for mapping vegetated areas with the required accuracy has proven to be challenging. One of the main issues is to estimate the trajectory for 3D model generation due to the GNSS signal obstructions, resulting in less accurate position estimation. An alternative is using techniques based on point cloud registration to ensure precise positioning (Liang et al., 2014; Zhou et al., 2024). However, another problem emerges in both forest and agricultural environments due to the similarity of features in the scene. In agricultural orchards, there are further challenges in this regard. Unlike boreal forests, where tree trunks can provide stable reference points (Liang; Hyyppä, 2013; Zhou et al., 2024), some orchards present repetitive patterns, such as rows of crops or plants with highly similar characteristics. Most of the identifiable objects in perennial crops consist of 18 Castanheiro, L. F. leaves or branches, which are not uniform and stable, due to movements caused by wind. These issues directly affect keypoint detection and matching for point cloud registration, and the resulting platform positioning, hindering the use of a BLS in everyday farming operations with the current technologies. Approaches have been explored to address these challenges, either using artificial targets in the environments (Jiang et al., 2022) or adapting existing algorithms for agricultural environments (Wang et al., 2021). However, the unique characteristics of tree types, such as canopy density and crown pattern, can provide different challenges and affect the performance of laser point cloud processing algorithms presented in the literature (Berk et al., 2020). For instance, perennial crops (e.g., orange, lemon, and coffee) present repetitive pattern, characteristics that are considered in this work. This work presents the detailed development steps of a BLS system designed for agricultural applications, which can also be applied in forests and other environments, including indoor areas. This thesis focuses on addressing the challenges of employing the BLS in perennial crops (e.g., orange orchards), which have repetitive features. Therefore, an algorithm for processing BLS data focusing on BLS trajectory estimation and point cloud registration to generate the full map of the environment is proposed. 1.2 MOTIVATION MLS enables obtaining measurements and a complete analysis of the structure of forests and crops that would be impractical or impossible to obtain through conventional methods. The MLS's ability to generate a detailed 3D model allows for extracting complex structural parameters, such as biomass estimation and canopy architecture. Therefore, the development of MLS has been presented in several journals, conferences, and the International Society of Photogrammetry and Remote Sensing (ISPRS - Working Group I/2). Despite laser scanning system capabilities, implementing BLS in farming environments faces some challenges, which motivated this work. First, the high cost of commercial laser scanning systems has been a significant barrier to their widespread use. Recent developments in cost-effective and lightweight sensors, such as the Ouster laser scanning system, make this technology more accessible for portable platforms. A second challenging issue in BLS is the estimation of the position and orientation of the platform without depending on the Global Navigation Satellite System (GNSS) data since the GNSS signals can 19 Castanheiro, L. F. be partially obstructed or blocked in vegetated areas. Another critical challenge, which also motivated this research, is data processing. The agricultural environment presents diverse and complex scenarios, each requiring specialized solutions for effective data processing and analysis. and pose different challenges that still need suitable solutions. In this work, an approach is proposed to address the challenges of employing a BLS for mapping perennial crops, especially focusing on orange orchards, an important commodity in Brazil, one of the world's leading orange producers. The solutions for assembling the hardware, processing raw data from an Ouster OS0-128 laser scanner, and producing a complete point cloud in such challenging environments ensure the originality of this work. 1.3 OBJECTIVES The main goals of this work were to develop and evaluate a BLS and to propose an approach to process the BLS data generating a 3D mapping of perennial crops, focusing on orange orchard. The specific objectives are: (1) to design and to integrate the components and sensors in a backpack-mounted platform; (2) to explore the use of the developed BLS for close- range applications, such as agriculture and forest mapping; (3) to analyse and assess approaches to estimate trajectory based solely on IMU measurements; (4) to assess feature extraction and matching techniques for 3D point collected in an orange orchard; (5) to develop a technique to generate point clouds and the incremental estimation of the trajectory with suitable accuracy. 1.4 THESIS STRUCTURE AND RELATED PUBLICATIONS Chapter 2 summarizes the studies performed by Castanheiro et al. (2022). A new backpack-mounted platform composed of a laser scanning system is introduced. A technical description of the proposed BLS, which includes a detailed description of the Ouster OS0-128 sensor, is presented. Then, the BLS feasibility study for a forestry urban environment is performed using WebSLAM, an Ouster solution for point cloud generation. Chapter 3 is based on the work presented by Castanheiro et al. (2024b), which assessed Ouster’s internal IMU data. Experiments were performed to verify better techniques to estimate platform orientation and position based solely on IMU data. Therefore, three algorithms freely available in GitHub were assessed. This chapter also presented a detailed analysis of Ouster's internal IMU and the discussion of estimation trajectory based on IMU, which is a challenging task. 20 Castanheiro, L. F. Chapter 4 presents an upgraded version of the developed BLS for agricultural environment mapping. This chapter also presents a study of point cloud registration methods in orange orchards. The experiments consisted of comparing feature-based and direct point cloud registration, and discussing their use in featureless environments, such as an orange orchard. This Chapter extends the study presented by Castanheiro et al. (2023). Chapter 5 presents an incremental registration technique to estimate the platform’s positions and orientation and to generate the 3D mapping of the environment. The mathematical model consisted of the point-to-plane approach to initially register individual point clouds collected by the BLS. A methodology for BLS data acquisition and processing is based on techniques reported in Chapters 2 to 4. The summary of the results and conclusions are presented in Chapter 6. 1.5 RELATED GRANTS AND INTERNATIONAL PARTNERSHIPS This doctoral research is part of a thematic project, funded by FAPESP, and aligned with the United Nations Sustainable Development Goals, specifically focusing on digital agriculture (“High resolution remote sensing for digital agriculture”), and a Capes-PrInt project (“From Precision Agriculture to Prescription Agriculture: insertion of Unesp in the context of Digital Agriculture and Bioeconomy”). The CAPES-PrInt project funded a mobility period of one year at the Finnish Geospatial Research Institute in the National Land Survey (FGI-NLS). This collaborative work significantly contributed to the development of the proposed solution. 96 Castanheiro, L. F. CHAPTER 6 FINAL REMARKS In this work, we proposed, developed and experimentally assessed a BLS with an Ouster OS0-128 laser scanner for agricultural applications. Performance assessments of this system in a forestry fragment and an orange orchard, two challenging and lack-explored environments, were presented. This final chapter discusses the open problems for future works based on the achievements. In Chapter 2, the performance of object reconstruction of the Ouster OS0-128 laser scanner was evaluated in an urban forest fragment using WebSLAM. The Ouster’s wide FoV provided a complete mapping of the environments, including tall trees (> 2 m) and urban objects with a high level of detail. This makes the system suitable for orange orchards where there are narrow spaces between tree lines and the trees are tall (~3 m). However, the 3D reconstruction with WebSLAM was not successful for back and forward paths. This problem can be explained by insufficient or wrong keypoint matches or drift in the estimated trajectory due to a very sharp trajectory turn. These findings show the need for improvements in navigation processing. To address these challenges, this research focuses on navigation processing to improve trajectory estimation (Chapter 3), the investigation of keypoints and features extraction techniques (Chapter 4), and the implementation of an approach to correctly generate a 3D model (Chapter 5). Focusing on the navigation system used, Chapter 3 presented the evaluation of the Ouster OS0-128’s internal IMU. The analysis focused on orientation estimation, exploring both Euler and quaternions for spatial representation. Three techniques were implemented and compared to assess the IMU performance in estimating short trajectories. Besides MEMS IMU having a high level of errors and noise, the experiments presented in that chapter showed that the use of an estimation filter, such as the Kalman filter, can improve the estimation. Therefore, DCM technique was used to estimate the initial values of the BLS trajectory. Chapter 4 presented an evaluation of keypoints detection, feature extraction and matching techniques for the registration of point clouds obtained in orange orchards. Through this assessment, key challenges were identified. The most significant limitation is due to the nature of the orange tree's canopy structures that make the environment featureless, where point clouds are mostly composed of laser pulses returned from leaves and ground. This study showed the limitations of classical point cloud registrations, which are, generally, developed and tested 97 Castanheiro, L. F. in environments with rich features (e.g., corners, lines, pole, etc.). The results highlighted the need for other approaches to handle the challenges of agricultural environments, leading to the implementation presented in Chapter 5. Chapter 5 presented an approach for incremental registration to evaluate the future potential of the proposed BLS in real-time applications. In this work, the use of planar features for point cloud registration, the estimation of the trajectory, and the generation of 3D mapping were assessed. The point-to-plane approach was employed since laser pulses returned from the ground populate most of the point clouds. The estimated trajectory maintained close alignment with the GNSS reference trajectory for the first 100 m of the path. However, after that, a deviation occurred, achieving an RMSE of 3.86 m and 7.19 m in planimetric and altimetric coordinates, respectively, at the end of the data acquisition path. Therefore, many improvements are still required to achieve higher accuracy for a longer trajectory, with refinement in the model or even the combination with other features (e.g., linear elements and full canopy structures) and sensors. Future works should focus on processing improvements in the trajectory estimation and point cloud generation. Developments focus on implementing a global adjustment, using GNSS data when available, and developing scan-to-map approaches. The use of the positions and orientation from IMU data as constraints in the LSM and implementing simultaneous adjustment of all observations could also improve the trajectory estimation. Automatic outlier detection, which remains a complex task in forest and agricultural scenarios, can be more explored. In this work, only the analysis of the residual vector of the LSM was performed. Other techniques can be implemented. In addition, the detection and exclusion of keyframes, not addressed in this work, could optimize the trajectory estimation process. The time processing deserves an assessment to determine an optimal number of matches versus processing time. The BLS performance evaluation should also be considered in future works. The assessment of the BLS and the proposed approach must be performed in areas with turns and varying the velocity since those present challenges for trajectory estimation and mapping. A comparative assessment with a higher-quality navigation system should also be performed. For agricultural applications, evaluating the point cloud reconstruction in areas with higher fruit density would assess a further application of the BLS for fruit detection and counting. The proposed BLS can be expanded beyond the current feasibility assessment study performed on an orange orchard. Experimental assessment in other agricultural crops (e.g., coffee) and forest environments are recommended. This assessment would validate the system's 98 Castanheiro, L. F. versatility. The integration of other sensors, such as dual-fisheye systems, multispectral cameras and other Ouster laser scanner, can be integrated into the BLS, resulting in many possibilities for multi-sensor systems and data processing methodologies. 99 Castanheiro, L. F. REFERENCES AGUIAR, A. S.; DOS SANTOS, F. N.; CUNHA, J. B.; SOBREIRA, H.; SOUSA, A. J. Localization and Mapping for Robots in Agriculture and Forestry: A Survey. Robotics, v. 9, n. 4, p. 97, 2020. https://doi.org/10.3390/robotics9040097. ALDOMA, A.; MARTON, Z.-C.; TOMBARI, F.; WOHLKINGER, W.; POTTHAST, C.; ZEISL, B.; RUSU, R. B.; GEDIKLI, S.; VINCZE, M. Tutorial: Point Cloud Library: Three- Dimensional Object Recognition and 6 DOF Pose Estimation. IEEE Robotics & Automation Magazine, v. 19, n. 3, p. 80–91, 2012. https://doi.org/10.1109/MRA.2012.2206675. ALSADIK, B. Ideal Angular Orientation of Selected 64-Channel Multi Beam Lidars for Mobile Mapping Systems. Remote Sensing, v. 12, n. 3, p. 510, 2020. https://doi.org/10.3390/rs12030510. ALSADIK, B.; REMONDINO, F. Flight Planning for LiDAR-Based UAS Mapping Applications. ISPRS International Journal of Geo-Information, v. 9, n. 6, p. 378, 2020. https://doi.org/10.3390/ijgi9060378. ARASTOUNIA, M.; LICHTI, D. D. Simultaneous identification, modeling and registration refinement of poles using laser scanning point clouds. ISPRS Journal of Photogrammetry and Remote Sensing, v. 181, p. 327–344, 2021. https://doi.org/10.1016/j.isprsjprs.2021.09.017. ARYAN, A.; BOSCHÉ, F.; TANG, P. Planning for terrestrial laser scanning in construction: A review. Automation in Construction, v. 125, p. 103551, 2021. https://doi.org/10.1016/j.autcon.2021.103551. BERK, P.; STAJNKO, D.; BELSAK, A.; HOCEVAR, M. Digital evaluation of leaf area of an individual tree canopy in the apple orchard using the LIDAR measurement system. Computers and Electronics in Agriculture, v. 169, p. 105158, 2020. https://doi.org/10.1016/j.compag.2019.105158. BERNREITER, L.; OTT, L.; NIETO, J.; SIEGWART, R.; CADENA, C. PHASER: A Robust and Correspondence-Free Global Pointcloud Registration. IEEE Robotics and Automation Letters, v. 6, n. 2, p. 855–862, 2021. https://doi.org/10.1109/LRA.2021.3052418. BESL, P. J.; MCKAY, N. D. Method for registration of 3-D shapes. (P. S. Schenker, Ed.). Em: ROBOTICS - DL TENTATIVE. Boston, MA: 1992. https://doi.org/10.1117/12.57955. BLASER, S.; NEBIKER, S.; WISLER, D. PORTABLE IMAGE-BASED HIGH PERFORMANCE MOBILE MAPPING SYSTEM IN UNDERGROUND ENVIRONMENTS – SYSTEM CONFIGURATION AND PERFORMANCE EVALUATION. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. IV-2/W5, p. 255–262, 2019. https://doi.org/10.5194/isprs-annals-IV-2-W5-255-2019. BURMEISTER, J.-M.; RICHTER, R.; REDER, S.; MUND, J.-P.; DÖLLNER, J. Tree Instance Segmentation in Urban 3D Point Clouds Using a Coarse-to-Fine Algorithm Based on Semantic Segmentation. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial 100 Castanheiro, L. F. Information Sciences, v. X-4-W5-2024, p. 79–86, 2024. https://doi.org/10.5194/isprs-annals- X-4-W5-2024-79-2024. CADENA, C.; CARLONE, L.; CARRILLO, H.; LATIF, Y.; SCARAMUZZA, D.; NEIRA, J.; REID, I.; LEONARD, J. J. Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age. IEEE Transactions on Robotics, v. 32, n. 6, p. 1309–1332, 2016. https://doi.org/10.1109/TRO.2016.2624754. CAMPOS, C.; ELVIRA, R.; RODRÍGUEZ, J. J. G.; M. MONTIEL, J. M.; D. TARDÓS, J. ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual–Inertial, and Multimap SLAM. IEEE Transactions on Robotics, v. 37, n. 6, p. 1874–1890, 2021a. https://doi.org/10.1109/TRO.2021.3075644. CAMPOS, M. B.; CASTANHEIRO, L. F.; SHAH, D.; WANG, Y.; KUKKO, A.; PUTTONEN, E. Overview and Benchmark on Multi-Modal Lidar Point Cloud Registration for Forest Applications. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. XLVIII-1–2024, p. 43–50, 2024. https://doi.org/10.5194/isprs-archives-XLVIII-1-2024-43-2024. CAMPOS, M. B.; LITKEY, P.; WANG, Y.; CHEN, Y.; HYYTI, H.; HYYPPÄ, J.; PUTTONEN, E. A Long-Term Terrestrial Laser Scanning Measurement Station to Continuously Monitor Structural and Phenological Dynamics of Boreal Forest Canopy. Frontiers in Plant Science, v. 11, p. 606752, 2021b. https://doi.org/10.3389/fpls.2020.606752. CAMPOS, M. B.; TOMMASELLI, A. M. G.; HONKAVAARA, E.; PROL, F. D. S.; KAARTINEN, H.; EL ISSAOUI, A.; HAKALA, T. A Backpack-Mounted Omnidirectional Camera with Off-the-Shelf Navigation Sensors for Mobile Terrestrial Mapping: Development and Forest Application. Sensors, v. 18, n. 3, p. 827, 2018. https://doi.org/10.3390/s18030827. CASTANHEIRO, L. F.; SILVA, M. F. DA; SILVA, R. M. DA; SANTOS, R. C. DOS; TOMMASELLI, A. M. G.; GALO, M. AVALIAÇÃO DE UM SISTEMA DE VARREDURA A LASER MÓVEL PARA ESTIMAÇÃO DO DAP EM EUCALIPTOS. Anais do Seminário Nacional de Fotogrametria e Sensoriamento Remoto no Ambiente Semiárido. Anais...2024a. CASTANHEIRO, L. F.; TOMMASELLI, A. M. G.; GARCIA, T. A. C.; CAMPOS, M. B.; KUKKO, A. POINT CLOUD REGISTRATION USING LASER DATA FROM AN ORANGE ORCHARD. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. XLVIII-1-W1-2023, p. 71–77, 2023. https://doi.org/10.5194/isprs-archives-XLVIII-1-W1-2023-71-2023. CASTANHEIRO, L. F.; TOMMASELLI, A. M. G.; HYYTI, H.; KUKKO, A. Experimental assessment of IMU data processing techniques for a backpack mobile laser scanning system. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. XLVIII-4–2024, 2024b. CASTANHEIRO, L. F.; TOMMASELLI, A. M. G.; MACHADO, M. V.; SANTOS, G. H.; NORBERTO, I. S.; REIS, T. T. THE USE OF A WIDE FOV LASER SCANNING SYSTEM AND A SLAM ALGORITHM FOR MOBILE APPLICATIONS. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. XLIII-B1- 2022, p. 181–187, 2022. https://doi.org/10.5194/isprs-archives-XLIII-B1-2022-181-2022. 101 Castanheiro, L. F. CHEN, H.; BHANU, B. 3D free-form object recognition in range images using local surface patches. Pattern Recognition Letters, v. 28, n. 10, p. 1252–1262, 2007. https://doi.org/10.1016/j.patrec.2007.02.009. CHEN, Y.; MEDIONI, G. Object modelling by registration of multiple range images. Image and Vision Computing, Range Image Understanding. v. 10, n. 3, p. 145–155, 1992. https://doi.org/10.1016/0262-8856(92)90066-C. CHENG, L.; CHEN, S.; LIU, X.; XU, H.; WU, Y.; LI, M.; CHEN, Y. Registration of Laser Scanning Point Clouds: A Review. Sensors, v. 18, n. 5, p. 1641, 2018. https://doi.org/10.3390/s18051641. CHO, H.; KIM, E. K.; KIM, S. Indoor SLAM application using geometric and ICP matching methods based on line features. Robotics and Autonomous Systems, v. 100, p. 206–224, 2018. https://doi.org/10.1016/j.robot.2017.11.011. DAI, W.; KAN, H.; TAN, R.; YANG, B.; GUAN, Q.; ZHU, N.; XIAO, W.; DONG, Z. Multisource forest point cloud registration with semantic-guided keypoints and robust RANSAC mechanisms. International Journal of Applied Earth Observation and Geoinformation, v. 115, p. 103105, 2022. https://doi.org/10.1016/j.jag.2022.103105. DAM, E. B.; KOCH, M.; LILLHOLM, M. Quaternions, Interpolation and Animation. Copenhagen, Denmark: Citeseer, 1998. v. 2 DAVENPORT, I. J.; HOLDEN, N.; GURNEY, R. J. Characterizing errors in airborne laser altimetry data to extract soil roughness. IEEE Transactions on Geoscience and Remote Sensing, v. 42, n. 10, p. 2130–2141, 2004. https://doi.org/10.1109/TGRS.2004.834648. DURRANT-WHYTE, H.; BAILEY, T. Simultaneous localization and mapping: part I. IEEE Robotics & Automation Magazine, v. 13, n. 2, p. 99–110, 2006. https://doi.org/10.1109/MRA.2006.1638022. EL ISSAOUI, A.; FENG, Z.; LEHTOMÄKI, M.; HYYPPÄ, E.; HYYPPÄ, H.; KAARTINEN, H.; KUKKO, A.; HYYPPÄ, J. Feasibility of Mobile Laser Scanning towards Operational Accurate Road Rut Depth Measurements. Sensors, v. 21, n. 4, p. 1180, 2021. https://doi.org/10.3390/s21041180. EL-SHEIMY, N.; VALEO, C.; HABIB, A. Digital terrain modeling: acquisition, manipulation, and applications. (No Title), 2005. FAITLI, T.; HAKALA, T.; KAARTINEN, H.; HYYPPÄ, J.; KUKKO, A. REAL-TIME LIDAR-INERTIAL POSITIONING AND MAPPING FOR FORESTRY AUTOMATION. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. XLVIII-1/W1-2023, p. 145–150, 2023. https://doi.org/10.5194/isprs- archives-XLVIII-1-W1-2023-145-2023. FLINT, A.; DICK, A.; HENGEL, A. VAN DEN. Thrift: Local 3D Structure Recognition. 9th Biennial Conference of the Australian Pattern Recognition Society on Digital Image Computing Techniques and Applications (DICTA 2007). Anais... Em: 9TH BIENNIAL CONFERENCE OF THE AUSTRALIAN PATTERN RECOGNITION SOCIETY ON DIGITAL IMAGE 102 Castanheiro, L. F. COMPUTING TECHNIQUES AND APPLICATIONS (DICTA 2007). 2007. https://doi.org/10.1109/DICTA.2007.4426794. FROME, A.; HUBER, D.; KOLLURI, R.; BÜLOW, T.; MALIK, J. Recognizing Objects in Range Data Using Regional Point Descriptors. (T. Pajdla, J. Matas, Eds.)Computer Vision - ECCV 2004. Anais...Berlin, Heidelberg: Springer, 2004. https://doi.org/10.1007/978-3-540- 24672-5_18. GAO, J.; SHA, J.; LI, H.; WANG, Y. A Robust and Fast GNSS-Inertial-LiDAR Odometry With INS-Centric Multiple Modalities by IESKF. IEEE Transactions on Instrumentation and Measurement, v. 73, p. 1–12, 2024. https://doi.org/10.1109/TIM.2024.3351253. GARCIA, T. A. C.; TOMMASELLI, A. M. G.; CASTANHEIRO, L. F.; CAMPOS, M. B. A photogrammetric approach for real‐time visual SLAM applied to an omnidirectional system. The Photogrammetric Record, v. 39, n. 187, p. 577–599, 2024. https://doi.org/10.1111/phor.12494. GHORBANI, F.; EBADI, H.; PFEIFER, N.; SEDAGHAT, A. Uniform and Competency-Based 3D Keypoint Detection for Coarse Registration of Point Clouds with Homogeneous Structure. Remote Sensing, v. 14, n. 16, p. 4099, 2022. https://doi.org/10.3390/rs14164099. GLENNIE, C. L.; HARTZELL, P. J. ACCURACY ASSESSMENT AND CALIBRATION OF LOW-COST AUTONOMOUS LIDAR SENSORS. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. XLIII-B1-2020, p. 371–376, 2020. https://doi.org/10.5194/isprs-archives-XLIII-B1-2020-371-2020. GRANT, D.; BETHEL, J.; CRAWFORD, M. Point-to-plane registration of terrestrial laser scans. ISPRS Journal of Photogrammetry and Remote Sensing, v. 72, p. 16–26, 2012. https://doi.org/10.1016/j.isprsjprs.2012.05.007. GUO, Y.; BENNAMOUN, M.; SOHEL, F.; LU, M.; WAN, J.; KWOK, N. M. A Comprehensive Performance Evaluation of 3D Local Feature Descriptors. International Journal of Computer Vision, v. 116, n. 1, p. 66–89, 2016. https://doi.org/10.1007/s11263- 015-0824-y. GUO, Y.; SOHEL, F.; BENNAMOUN, M.; LU, M.; WAN, J. TriSI: A Distinctive Local Surface Descriptor for 3D Modeling and Object Recognition. Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information Visualization Theory and Applications - Volume 1: GRAPP, (VISIGRAPP 2013). Anais...SciTePress, 2013a. https://doi.org/10.5220/0004277600860093. GUO, Y.; SOHEL, F.; BENNAMOUN, M.; LU, M.; WAN, J. Rotational Projection Statistics for 3D Local Surface Description and Object Recognition. International Journal of Computer Vision, v. 105, n. 1, p. 63–86, 2013b. https://doi.org/10.1007/s11263-013-0627-y. HALL, B. C. Lie Groups, Lie Algebras, and Representations. Em: HALL, B. C. (Ed.). Quantum Theory for Mathematicians. New York, NY: Springer, 2013. p. 333–366. https://doi.org/10.1007/978-1-4614-7116-5_16. 103 Castanheiro, L. F. HARRIS, C.; STEPHENS, M. A combined corner and edge detector. Alvey vision conference. Anais...Citeseer, 1988. HYYPPÄ, E.; MUHOJOKI, J.; YU, X.; KUKKO, A.; KAARTINEN, H.; HYYPPÄ, J. Efficient coarse registration method using translation- and rotation-invariant local descriptors towards fully automated forest inventory. ISPRS Open Journal of Photogrammetry and Remote Sensing, v. 2, p. 100007, 2021. https://doi.org/10.1016/j.ophoto.2021.100007. HYYTI, H.; VISALA, A. A DCM Based Attitude Estimation Algorithm for Low-Cost MEMS IMUs. International Journal of Navigation and Observation, v. 2015, n. 1, p. 503814, 2015. https://doi.org/10.1155/2015/503814. INVENSENSE, INC. TDK InvenSense: ICM-20948 World’s Lowest Power 9-Axis MEMS MotionTrackingTM Device. , 2021. Disponível em: JIANG, S.; QI, P.; HAN, L.; LIU, L.; LI, Y.; HUANG, Z.; LIU, Y.; HE, X. Navigation system for orchard spraying robot based on 3D LiDAR SLAM with NDT_ICP point cloud registration. Computers and Electronics in Agriculture, v. 220, p. 108870, 2024. https://doi.org/10.1016/j.compag.2024.108870. JIANG, S.; WANG, S.; YI, Z.; ZHANG, M.; LV, X. Autonomous Navigation System of Greenhouse Mobile Robot Based on 3D Lidar and 2D Lidar SLAM. Frontiers in Plant Science, v. 13, 2022. https://doi.org/10.3389/fpls.2022.815218. JIN, S.; SU, Y.; WU, F.; PANG, S.; GAO, S.; HU, T.; LIU, J.; GUO, Q. Stem–Leaf Segmentation and Phenotypic Trait Extraction of Individual Maize Using Terrestrial LiDAR Data. IEEE Transactions on Geoscience and Remote Sensing, v. 57, n. 3, p. 1336–1346, 2019. https://doi.org/10.1109/TGRS.2018.2866056. JOHNSON, A. E.; HEBERT, M. Using spin images for efficient object recognition in cluttered 3D scenes. IEEE Transactions on Pattern Analysis and Machine Intelligence, v. 21, n. 5, p. 433–449, 1999. https://doi.org/10.1109/34.765655. JÚNIOR, G. P. C.; REZENDE, A. M. C.; MIRANDA, V. R. F.; FERNANDES, R.; AZPÚRUA, H.; NETO, A. A.; PESSIN, G.; FREITAS, G. M. EKF-LOAM: An Adaptive Fusion of LiDAR SLAM With Wheel Odometry and Inertial Data for Confined Spaces With Few Geometric Features. IEEE Transactions on Automation Science and Engineering, v. 19, n. 3, p. 1458–1471, 2022. https://doi.org/10.1109/TASE.2022.3169442. KAARTINEN, H.; HYYPPÄ, J.; VASTARANTA, M.; KUKKO, A.; JAAKKOLA, A.; YU, X.; PYÖRÄLÄ, J.; LIANG, X.; LIU, J.; WANG, Y.; KAIJALUOTO, R.; MELKAS, T.; HOLOPAINEN, M.; HYYPPÄ, H. Accuracy of Kinematic Positioning Using Global Satellite Navigation Systems under Forest Canopies. Forests, v. 6, n. 9, p. 3218–3236, 2015. https://doi.org/10.3390/f6093218. KIM, B.; JUNG, C.; SHIM, D. H.; AGHA–MOHAMMADI, A. Adaptive Keyframe Generation based LiDAR Inertial Odometry for Complex Underground Environments. 2023 IEEE International Conference on Robotics and Automation (ICRA). Anais... Em: 2023 IEEE 104 Castanheiro, L. F. INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA). London, United Kingdom: IEEE, 2023. https://doi.org/10.1109/ICRA48891.2023.10161207. KOENIG, K.; HÖFLE, B.; HÄMMERLE, M.; JARMER, T.; SIEGMANN, B.; LILIENTHAL, H. Comparative classification analysis of post-harvest growth detection from terrestrial LiDAR point clouds in precision agriculture. ISPRS Journal of Photogrammetry and Remote Sensing, v. 104, p. 112–125, 2015. https://doi.org/10.1016/j.isprsjprs.2015.03.003. KUKKO, A.; KAARTINEN, H.; HYYPPÄ, J.; CHEN, Y. Multiplatform Mobile Laser Scanning: Usability and Performance. Sensors, v. 12, n. 9, p. 11712–11733, 2012. https://doi.org/10.3390/s120911712. KUKKO, A.; KAIJALUOTO, R.; KAARTINEN, H.; LEHTOLA, V. V.; JAAKKOLA, A.; HYYPPÄ, J. Graph SLAM correction for single scanner MLS forest data under boreal forest canopy. ISPRS Journal of Photogrammetry and Remote Sensing, v. 132, p. 199–209, 2017. https://doi.org/10.1016/j.isprsjprs.2017.09.006. LEHTOLA, V. V.; KAARTINEN, H.; NÜCHTER, A.; KAIJALUOTO, R.; KUKKO, A.; LITKEY, P.; HONKAVAARA, E.; ROSNELL, T.; VAAJA, M. T.; VIRTANEN, J.-P.; KURKELA, M.; EL ISSAOUI, A.; ZHU, L.; JAAKKOLA, A.; HYYPPÄ, J. Comparison of the Selected State-Of-The-Art 3D Indoor Scanning and Point Cloud Generation Methods. Remote Sensing, v. 9, n. 8, p. 796, 2017. https://doi.org/10.3390/rs9080796. LEHTOMÄKI, M.; KUKKO, A.; MATIKAINEN, L.; HYYPPÄ, J.; KAARTINEN, H.; JAAKKOLA, A. Power line mapping technique using all-terrain mobile laser scanning. Automation in Construction, v. 105, p. 102802, 2019. https://doi.org/10.1016/j.autcon.2019.03.023. LEVOY, M.; GERTH, J.; CURLESS, B.; PULL, K. The Stanford 3D scanning repository. , 2005. Disponível em: LI, L.; WANG, R.; ZHANG, X. A Tutorial Review on Point Cloud Registrations: Principle, Classification, Comparison, and Technology Challenges. Mathematical Problems in Engineering, v. 2021, n. 1, p. 9953910, 2021. https://doi.org/10.1155/2021/9953910. LIANG, X.; HYYPPÄ, J. Automatic Stem Mapping by Merging Several Terrestrial Laser Scans at the Feature and Decision Levels. Sensors, v. 13, n. 2, p. 1614–1634, 2013. https://doi.org/10.3390/s130201614. LIANG, X.; KANKARE, V.; HYYPPÄ, J.; WANG, Y.; KUKKO, A.; HAGGRÉN, H.; YU, X.; KAARTINEN, H.; JAAKKOLA, A.; GUAN, F.; HOLOPAINEN, M.; VASTARANTA, M. Terrestrial laser scanning in forest inventories. ISPRS Journal of Photogrammetry and Remote Sensing, Theme issue “State-of-the-art in photogrammetry, remote sensing and spatial information science”. v. 115, p. 63–77, 2016. https://doi.org/10.1016/j.isprsjprs.2016.01.006. LIANG, X.; KUKKO, A.; KAARTINEN, H.; HYYPPÄ, J.; YU, X.; JAAKKOLA, A.; WANG, Y. Possibilities of a Personal Laser Scanning System for Forest Mapping and Ecosystem Services. Sensors, v. 14, n. 1, p. 1228–1248, 2014. https://doi.org/10.3390/s140101228. 105 Castanheiro, L. F. LIANG, X.; LITKEY, P.; HYYPPA, J.; KAARTINEN, H.; VASTARANTA, M.; HOLOPAINEN, M. Automatic Stem Mapping Using Single-Scan Terrestrial Laser Scanning. IEEE Transactions on Geoscience and Remote Sensing, v. 50, n. 2, p. 661–670, 2012. https://doi.org/10.1109/TGRS.2011.2161613. LIU, J.; LIANG, X.; HYYPPÄ, J.; YU, X.; LEHTOMÄKI, M.; PYÖRÄLÄ, J.; ZHU, L.; WANG, Y.; CHEN, R. Automated matching of multiple terrestrial laser scans for stem mapping without the use of artificial references. International Journal of Applied Earth Observation and Geoinformation, v. 56, p. 13–23, 2017. https://doi.org/10.1016/j.jag.2016.11.003. LOWE, T.; MOGHADAM, P.; EDWARDS, E.; WILLIAMS, J. Canopy density estimation in perennial horticulture crops using 3D spinning lidar SLAM. Journal of Field Robotics, v. 38, n. 4, p. 598–618, 2021. https://doi.org/10.1002/rob.22006. MADGWICK, S. O. H.; HARRISON, A. J. L.; VAIDYANATHAN, R. Estimation of IMU and MARG orientation using a gradient descent algorithm. 2011 IEEE International Conference on Rehabilitation Robotics. Anais... Em: 2011 IEEE INTERNATIONAL CONFERENCE ON REHABILITATION ROBOTICS. 2011. https://doi.org/10.1109/ICORR.2011.5975346. MAHONY, R.; HAMEL, T.; TRUMPF, J.; LAGEMAN, C. Nonlinear attitude observers on SO(3) for complementary and compatible measurements: A theoretical study. Proceedings of the 48h IEEE Conference on Decision and Control (CDC) held jointly with 2009 28th Chinese Control Conference. Anais... Em: PROCEEDINGS OF THE 48H IEEE CONFERENCE ON DECISION AND CONTROL (CDC) HELD JOINTLY WITH 2009 28TH CHINESE CONTROL CONFERENCE. 2009. https://doi.org/10.1109/CDC.2009.5399821. MAIMAITIJIANG, M.; SAGAN, V.; ERKBOL, H.; ADRIAN, J.; NEWCOMB, M.; LEBAUER, D.; PAULI, D.; SHAKOOR, N.; MOCKLER, T. C. UAV-BASED SORGHUM GROWTH MONITORING: A COMPARATIVE ANALYSIS OF LIDAR AND PHOTOGRAMMETRY. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. V-3–2020, p. 489–496, 2020. https://doi.org/10.5194/isprs- annals-V-3-2020-489-2020. MAO, W.; LIU, H.; HAO, W.; YANG, F.; LIU, Z. Development of a Combined Orchard Harvesting Robot Navigation System. Remote Sensing, v. 14, n. 3, p. 675, 2022. https://doi.org/10.3390/rs14030675. MIKHAIL, E. M.; ACKERMANN, F. E. Observations and least squares. 1. ed. New York: Dun-Donnelley, 1976. MISHRA, S.; PANDEY, G.; SARIPALLI, S. Extrinsic Calibration of a 3D-LIDAR and a Camera. 2020 IEEE Intelligent Vehicles Symposium (IV). Anais... Em: 2020 IEEE INTELLIGENT VEHICLES SYMPOSIUM (IV). 2020. https://doi.org/10.1109/IV47402.2020.9304750. MONICO, J. F. G. Posicionamento pelo GNSS: descrição, fundamentos e aplicações. [s.l.] Editora Unesp, 2007. MORADI, L.; SAADATSERESHT, M.; SHOKRZADEH, P. DEVELOPMENT OF A VOXEL BASED LOCAL PLANE FITTING FOR MULTI-SCALE REGISTRATION OF 106 Castanheiro, L. F. SEQUENTIAL MLS POINT CLOUDS. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. X-4/W1-2022, p. 523–530, 2023. https://doi.org/10.5194/isprs-annals-X-4-W1-2022-523-2023. MORIN, K. W. Calibration of airborne laser scanners. University of Calgary, , 2002. Disponível em: . Acesso em: 6 ago. 2024 MUR-ARTAL, R.; TARDOS, J. D. ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras. IEEE Transactions on Robotics, v. 33, n. 5, p. 1255–1262, 2017. https://doi.org/10.1109/TRO.2017.2705103. MYRONENKO, A.; SONG, X. Point Set Registration: Coherent Point Drift. IEEE Transactions on Pattern Analysis and Machine Intelligence, v. 32, n. 12, p. 2262–2275, 2010. https://doi.org/10.1109/TPAMI.2010.46. NGUYEN, H. L.; BELTON, D.; HELMHOLZ, P. COMPARATIVE STUDY OF AUTOMATIC PLANE FITTING REGISTRATION FOR MLS SPARSE POINT CLOUDS WITH DIFFERENT PLANE SEGMENTATION METHODS. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. IV-2/W4, p. 115– 122, 2017. https://doi.org/10.5194/isprs-annals-IV-2-W4-115-2017. ONDA, K.; OISHI, T.; KURODA, Y. Dynamic Environment Recognition for Autonomous Navigation with Wide FOV 3D-LiDAR⁎. IFAC-PapersOnLine, 12th IFAC Symposium on Robot Control SYROCO 2018. v. 51, n. 22, p. 530–535, 2018. https://doi.org/10.1016/j.ifacol.2018.11.579. OUSTER. Why Digital Lidar is the Future | Ouster. Disponível em: . Acesso em: 7 ago. 2024. OUSTER, INC. Firmware User Manual: Firmware v2.3.1 for all Ouster sensors. , 2022. Disponível em: . Acesso em: 7 ago. 2024 OUSTER, INC. Datasheet: OS0 Ultra-Wide View High-Resolution Imaging Lidar. , 2024a. Disponível em: . Acesso em: 1 jul. 2024 OUSTER, INC. Ouster Studio Quick Start Guide V2.0. , 2024b. Disponível em: . Acesso em: 7 ago. 2024 PACALA, A. How Multi-Beam Flash Lidar Works | Ouster. Disponível em: . Acesso em: 6 ago. 2024. PACALA, A. Ouster’s next step: 128 channel lidar sensors, long range, and an ultra-wide field of view | Ouster. Disponível em: . Acesso em: 7 ago. 2024. 107 Castanheiro, L. F. PACALA, A.; SHU, M. L. Solid-state electronic scanning laser array with high-side and low- side switches for increased channels. , 2023. Disponível em: . Acesso em: 7 ago. 2024 PAGLIAI, A.; AMMONIACI, M.; SARRI, D.; LISCI, R.; PERRIA, R.; VIERI, M.; D’ARCANGELO, M. E. M.; STORCHI, P.; KARTSIOTIS, S.-P. Comparison of Aerial and Ground 3D Point Clouds for Canopy Size Assessment in Precision Viticulture. Remote Sensing, v. 14, n. 5, p. 1145, 2022. https://doi.org/10.3390/rs14051145. PARK, S.-Y.; SUBBARAO, M. An accurate and fast point-to-plane registration technique. Pattern Recognition Letters, v. 24, n. 16, p. 2967–2976, 2003. https://doi.org/10.1016/S0167- 8655(03)00157-0. PENG, Y.; LIN, S.; WU, H.; CAO, G. Point Cloud Registration Based on Fast Point Feature Histogram Descriptors for 3D Reconstruction of Trees. Remote Sensing, v. 15, n. 15, p. 3775, 2023. https://doi.org/10.3390/rs15153775. PFEIFFER, S. A.; GUEVARA, J.; CHEEIN, F. A.; SANZ, R. Mechatronic terrestrial LiDAR for canopy porosity and crown surface estimation. Computers and Electronics in Agriculture, v. 146, p. 104–113, 2018. https://doi.org/10.1016/j.compag.2018.01.022. PITKÄNEN, J.; MALTAMO, M.; HYYPPÄ, J.; YU, X. ADAPTIVE METHODS FOR INDIVIDUAL TREE DETECTION ON AIRBORNE LASER BASED CANOPY HEIGHT MODEL. v. 36, n. 8, p. 187–191, 2004. RIVERA, G.; PORRAS, R.; FLORENCIA, R.; SÁNCHEZ-SOLÍS, J. P. LiDAR applications in precision agriculture for cultivating crops: A review of recent advances. Computers and Electronics in Agriculture, v. 207, p. 107737, 2023. https://doi.org/10.1016/j.compag.2023.107737. RUSU, R. B.; BLODOW, N.; BEETZ, M. Fast Point Feature Histograms (FPFH) for 3D registration. 2009 IEEE International Conference on Robotics and Automation. Anais... Em: 2009 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION. 2009. https://doi.org/10.1109/ROBOT.2009.5152473. RUSU, R. B.; BLODOW, N.; MARTON, Z. C.; BEETZ, M. Aligning point cloud views using persistent feature histograms. 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems. Anais... Em: 2008 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS. Nice: IEEE, 2008. https://doi.org/10.1109/IROS.2008.4650967. RUSU, R. B.; COUSINS, S. 3D is here: Point Cloud Library (PCL). 2011 IEEE International Conference on Robotics and Automation. Anais... Em: 2011 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION. 2011. https://doi.org/10.1109/ICRA.2011.5980567. SALTI, S.; TOMBARI, F.; DI STEFANO, L. SHOT: Unique signatures of histograms for surface and texture description. Computer Vision and Image Understanding, v. 125, p. 251– 264, 2014. https://doi.org/10.1016/j.cviu.2014.04.011. 108 Castanheiro, L. F. SALTI, S.; TOMBARI, F.; STEFANO, L. D. A Performance Evaluation of 3D Keypoint Detectors. Visualization and Transmission 2011 International Conference on 3D Imaging, Modeling, Processing. Anais... Em: VISUALIZATION AND TRANSMISSION 2011 INTERNATIONAL CONFERENCE ON 3D IMAGING, MODELING, PROCESSING. 2011. https://doi.org/10.1109/3DIMPVT.2011.37. SHAN, J.; TOTH, C. K. Topographic laser ranging and scanning: principles and processing. [s.l.] CRC press, 2008. SHAN, T.; WANG, J.; CHEN, F.; SZENHER, P.; ENGLOT, B. Simulation-based lidar super- resolution for ground vehicles. Robotics and Autonomous Systems, v. 134, p. 103647, 2020. https://doi.org/10.1016/j.robot.2020.103647. SHAO, J.; ZHANG, W.; LUO, L.; CAI, S.; JIANG, H. SLAM-BASED BACKPACK LASER SCANNING FOR FOREST PLOT MAPPING. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, v. V-2–2020, p. 267–271, 2020. https://doi.org/10.5194/isprs-annals-V-2-2020-267-2020. SHENDRYK, Y.; SOFONIA, J.; GARRARD, R.; RIST, Y.; SKOCAJ, D.; THORBURN, P. Fine-scale prediction of biomass and leaf nitrogen content in sugarcane using UAV LiDAR and multispectral imaging. International Journal of Applied Earth Observation and Geoinformation, v. 92, p. 102177, 2020. https://doi.org/10.1016/j.jag.2020.102177. SI, H.; QIU, J.; LI, Y. A Review of Point Cloud Registration Algorithms for Laser Scanners: Applications in Large-Scale Aircraft Measurement. Applied Sciences, v. 12, n. 20, p. 10247, 2022. https://doi.org/10.3390/app122010247. SILVA, M. F. DA; CASTANHEIRO, L. F.; TOMMASELLI, A. M. G.; DOS SANTOS, R. C.; GALO, M. DETECÇÃO DE ÁRVORES EM AMBIENTE URBANO A PARTIR DE DADOS LIDAR COLETADOS POR SISTEMA MÓVEL PORTÁTIL. Anais do XIII Colóquio Brasileiro de Ciências Geodésicas. Anais...XIII Colóquio Brasileiro de Ciências Geodésicas, 2024. https://doi.org/10.5281/ZENODO.14041226. SILVA, M. F. DA; SANTOS, R. C. DOS; GALO, M. Detection and segmentation of orange fruit in 3D point clouds generated by a terrestrial lidar system. 2005 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2005. Anais...IEEE, 2024. SINHA, K.; KELSEY, C.; FUHRMAN, J.; CHAUHAN, S.; KOCHHAR, I.; CARLSON, R.; MCGANN, F.; STOLZ, E.; LAMBERT, J.; FU, A.; AKOS, D. Functional LiDAR Analysis of Structural Health (FLASH). [s.l.] University of Colorado, Department of Aerospace Engineering Sciences (ASEN 4028), 2021. Disponível em: . SIPIRAN, I.; BUSTOS, B. Harris 3D: a robust extension of the Harris operator for interest point detection on 3D meshes. The Visual Computer, v. 27, n. 11, p. 963–976, 2011. https://doi.org/10.1007/s00371-011-0610-y. STANCELOVA, P.; SIKUDOVA, E.; CERNEKOVA, Z. Performance Evaluation of Selected 3D Keypoint Detector–Descriptor Combinations. Em: CHMIELEWSKI, L. J.; KOZERA, R.; ORŁOWSKI, A. (Eds.). Computer Vision and Graphics. Lecture Notes in Computer Science. 109 Castanheiro, L. F. Cham: Springer International Publishing, 2020. v. 12334p. 188–200. https://doi.org/10.1007/978-3-030-59006-2_17. STREIFF, D.; BERNREITER, L.; TSCHOPP, F.; FEHR, M.; SIEGWART, R. 3D3L: Deep Learned 3D Keypoint Detection and Description for LiDARs. 2021 IEEE International Conference on Robotics and Automation (ICRA). Anais... Em: 2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA). 2021. https://doi.org/10.1109/ICRA48506.2021.9560926. TAGLIABUE, A.; TORDESILLAS, J.; CAI, X.; SANTAMARIA-NAVARRO, A.; HOW, J. P.; CARLONE, L.; AGHA-MOHAMMADI, A. LION: Lidar-Inertial Observability-Aware Navigator for Vision-Denied Environments. Em: SICILIANO, B.; LASCHI, C.; KHATIB, O. (Eds.). Experimental Robotics. Springer Proceedings in Advanced Robotics. Cham: Springer International Publishing, 2021. v. 19p. 380–390. https://doi.org/10.1007/978-3-030-71151- 1_34. TOMBARI, F.; SALTI, S.; DI STEFANO, L. Unique Signatures of Histograms for Local Surface Description. (K. Daniilidis, P. Maragos, N. Paragios, Eds.)Computer Vision – ECCV 2010. Anais...Berlin, Heidelberg: Springer, 2010a. https://doi.org/10.1007/978-3-642-15558- 1_26. TOMBARI, F.; SALTI, S.; DI STEFANO, L. Unique shape context for 3d data description. Proceedings of the ACM workshop on 3D object retrieval. Anais... Em: MM ’10: ACM MULTIMEDIA CONFERENCE. Firenze Italy: ACM, 2010b. https://doi.org/10.1145/1877808.1877821. TSIOURVA, M.; PAPACHRISTOS, C. LiDAR Imaging-based Attentive Perception. 2020 International Conference on Unmanned Aircraft Systems (ICUAS). Anais... Em: 2020 INTERNATIONAL CONFERENCE ON UNMANNED AIRCRAFT SYSTEMS (ICUAS). 2020. https://doi.org/10.1109/ICUAS48674.2020.9213910. UMEYAMA, S. Least-squares estimation of transformation parameters between two point patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, v. 13, n. 4, p. 376–380, 1991. https://doi.org/10.1109/34.88573. UNITED NATIONS. Transforming our World: The 2030 Agenda for Sustainable Development. United Nations, , 2015. Disponível em: VANÍČEK, P. Introduction to adjustment calculus. [s.l.] Department of Surveying Engineering, University of New Brunswick, 1974. VEZETEU, E.; HYYTI, H.; KYRKI, V.; MANNINEN, P.; MAANPÄÄ, J.; TAHER, J.; HYYPPÄ, J. FRIS-LOAM: Fast Range Image Segmentation for LiDAR Odometry and Mapping in complex environments. 202-. https://github.com/eugeniu1994/FRIS-LOAM. VINCE, J. Quaternions for computer graphics. [s.l.] Springer, 2011. VOSSELMAN, G.; MAAS, H. G. (EDS.). Airborne and terrestrial laser scanning. [s.l.] CRC Press (Taylor & Francis), 2010. 110 Castanheiro, L. F. WALLACE, L. Assessing the stability of canopy maps produced from UAV-LiDAR data. 2013 IEEE International Geoscience and Remote Sensing Symposium - IGARSS. Anais... Em: 2013 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM - IGARSS. 2013. https://doi.org/10.1109/IGARSS.2013.6723679. WANG, K.; ZHOU, J.; ZHANG, W.; ZHANG, B. Mobile LiDAR Scanning System Combined with Canopy Morphology Extracting Methods for Tree Crown Parameters Evaluation in Orchards. Sensors, v. 21, n. 2, p. 339, 2021. https://doi.org/10.3390/s21020339. WANG, Z.; WU, Y.; NIU, Q. Multi-Sensor Fusion in Automated Driving: A Survey. IEEE Access, v. 8, p. 2847–2868, 2019. https://doi.org/10.1109/ACCESS.2019.2962554. WEHR, A.; LOHR, U. Airborne laser scanning—an introduction and overview. ISPRS Journal of Photogrammetry and Remote Sensing, v. 54, n. 2, p. 68–82, 1999. https://doi.org/10.1016/S0924-2716(99)00011-8. WON, S. P.; GOLNARAGHI, F. A Triaxial Accelerometer Calibration Method Using a Mathematical Model. IEEE Transactions on Instrumentation and Measurement, v. 59, n. 8, p. 2144–2153, 2010. https://doi.org/10.1109/TIM.2009.2031849. YANG, J.; XIAO, Y.; CAO, Z. Toward the Repeatability and Robustness of the Local Reference Frame for 3D Shape Matching: An Evaluation. IEEE Transactions on Image Processing, v. 27, n. 8, p. 3766–3781, 2018. https://doi.org/10.1109/TIP.2018.2827330. YE, H.; CHEN, Y.; LIU, M. Tightly Coupled 3D Lidar Inertial Odometry and Mapping. 2019 International Conference on Robotics and Automation (ICRA). Anais... Em: 2019 INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA). 2019. https://doi.org/10.1109/ICRA.2019.8793511. YIĞIT, A. Y.; GAMZE HAMAL, S. N.; ULVI, A.; YAKAR, M. Comparative analysis of mobile laser scanning and terrestrial laser scanning for the indoor mapping. Building Research & Information, v. 52, n. 4, p. 402–417, 2024. https://doi.org/10.1080/09613218.2023.2227900. ZHANG, J.; SINGH, S. LOAM: Lidar odometry and mapping in real-time. Robotics: Science and systems. Anais...Berkeley, CA, 2014. ZHANG, Y.; SHI, P.; LI, J. 3D LiDAR SLAM: A survey. The Photogrammetric Record, v. 39, n. 186, p. 457–517, 2024. https://doi.org/10.1111/phor.12497. ZHONG, Y. Intrinsic shape signatures: A shape descriptor for 3D object recognition. 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops. Anais... Em: 2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS, ICCV WORKSHOPS. 2009. https://doi.org/10.1109/ICCVW.2009.5457637. ZHOU, H.; ZHANG, J.; GE, L.; YU, X.; WANG, Y.; ZHANG, C. Research on volume prediction of single tree canopy based on three-dimensional (3D) LiDAR and clustering segmentation. International Journal of Remote Sensing, v. 42, n. 2, p. 738–755, 2021. https://doi.org/10.1080/01431161.2020.1811917. 111 Castanheiro, L. F. ZHOU, T.; ZHAO, C.; WINGREN, C. P.; FEI, S.; HABIB, A. Forest feature LiDAR SLAM (F2-LSLAM) for backpack systems. ISPRS Journal of Photogrammetry and Remote Sensing, v. 212, p. 96–121, 2024. https://doi.org/10.1016/j.isprsjprs.2024.04.025. RESSALVA - texto parcial.pdf castanheiro_lf_dr_prud.pdf