Bundle Block Adjustment of CBERS-2B HRC Imagery Combining Control Points and Lines

Abstract

The aim of this work is to present the results of the assessment of rigorous bundle block adjustment models for orbital imagery of the High-Resolution Camera (HRC) of the CBERS (China-Brazil Earth Resources Satellite). The work is focused on the experimental assessment of the combined use of ground control points (GCP) and ground control lines (GCL) in a block adjustment. The mathematical models relating object and image spaces are based on collinearity (for points) and coplanarity (for lines) conditions with polynomial modelling of the spacecraft trajectory and attitude. These models were implemented in the software TMS (triangulation with multiple sensors) with multi-feature control (GCPs and GCLs) developed in-house. Experiments on a block of four CBERS-2B HRC images were accomplished using both GCPs and GCLs. The results show that the combination of the collinearity and coplanarity models can provide better results in the bundle block adjustment process than conventional bundle adjustment with GCPs only. A systematic error in the inner geometry of HRC camera caused by the displacement of one of the three CCD sensors and the lack of proper correction when fusing the three images to generate level 1 images was also verified. Experiments to evaluate the effects of this systematic error are also presented.