Resposta espectral: análise de componentes principais para a diferenciação de arranjos espaciais na soja
Carregando...
Data
2023-03-03
Autores
Santos, Jamile do Nascimento
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O estudo com sensoriamento remoto na soja, embora importante concentra-se na estimativa ou na predição da produtividade. Entretanto, a produtividade da soja pode variar de acordo com a população de plantas e o espaçamento entre linhas. Dessa forma o objetivo deste estudo foi diferenciar arranjos espaciais de soja a partir de imagens multiespectrais por meio de índices de vegetação. Foram utilizados dois espaçamentos de semeadura: convencional (C) com 0,45 m e linhas duplas intercaladas (D) com duas linhas de 0,30 m por uma de 0,60 m. Além disso, foram adotadas três populações de plantas: P1= 300, P2= 350 e P3= 400 mil plantas ha-1. As imagens multiespectrais foram coletadas aos 45, 55, 65,75, 85 e 95 dias após a emergência, os parâmetros avaliados foram: índices de vegetação, bandas espectrais, número de vagens por planta, número de vagens por grão, massa de mil grãos. A análise de componentes principais foi aplicada para diferenciar os arranjos espaciais de soja. Os resultados mostraram que a banda do vermelho e do infravermelho próximo são as bandas espectrais mais responsivas para diferenciação dos arranjos dos 45 aos 75 dias após a emergência. Independentemente dos dias após emergência, o Índice de Vegetação por Diferença Normalizada (NDVI), Índice de Vegetação Ajustado ao Solo (SAVI), Índice De Vegetação Não-Linear (NLI) apresentaram a mesma capacidade de diferenciar os arranjos. Essas descobertas fornecem mais informações sobre a obtenção remota da informação do arranjo espacial que auxiliará na tomada de decisão de qual arranjo usar.
The study on remote sensing in soybean cultivation, although important, focuses primarily on estimating or predicting productivity. However, soybean productivity can vary based on plant population and row spacing. Therefore, the objective of this study was to differentiate soybean spatial arrangements using multispectral imagery and vegetation indices. Two seeding spacings were used: conventional (C) with 0.45m between rows, and intercalated double rows (D) with two rows spaced at 0.30m and one row at 0.60m. Additionally, three plant populations were adopted: P1=300, P2=350, and P3=400 thousand plants per hectare. Multispectral images were collected at 45, 55, 65, 75, 85, and 95 days after emergence. The evaluated parameters included vegetation indices, spectral bands, number of pods per plant, number of pods per grain, and thousand grain weight. Principal component analysis was applied to differentiate the soybean spatial arrangements. The results showed that the red and near-infrared spectral bands were the most responsive for differentiating the arrangements from 45 to 75 days after emergence. Regardless of the days after emergence, the Normalized Difference Vegetation Index (NDVI), SoilAdjusted Vegetation Index (SAVI), and Non-Linear Vegetation Index (NLI) exhibited the same capability to differentiate the arrangements. These findings provide further information on remotely obtaining spatial arrangement information, which will assist in decision-making regarding which arrangement to use.
The study on remote sensing in soybean cultivation, although important, focuses primarily on estimating or predicting productivity. However, soybean productivity can vary based on plant population and row spacing. Therefore, the objective of this study was to differentiate soybean spatial arrangements using multispectral imagery and vegetation indices. Two seeding spacings were used: conventional (C) with 0.45m between rows, and intercalated double rows (D) with two rows spaced at 0.30m and one row at 0.60m. Additionally, three plant populations were adopted: P1=300, P2=350, and P3=400 thousand plants per hectare. Multispectral images were collected at 45, 55, 65, 75, 85, and 95 days after emergence. The evaluated parameters included vegetation indices, spectral bands, number of pods per plant, number of pods per grain, and thousand grain weight. Principal component analysis was applied to differentiate the soybean spatial arrangements. The results showed that the red and near-infrared spectral bands were the most responsive for differentiating the arrangements from 45 to 75 days after emergence. Regardless of the days after emergence, the Normalized Difference Vegetation Index (NDVI), SoilAdjusted Vegetation Index (SAVI), and Non-Linear Vegetation Index (NLI) exhibited the same capability to differentiate the arrangements. These findings provide further information on remotely obtaining spatial arrangement information, which will assist in decision-making regarding which arrangement to use.
Descrição
Palavras-chave
Análise multivariada, Glycine max (L.) Merr, População de plantas, Sensoriamento remoto, Soja