Above-Ground Biomass Estimation of Peanut by Integrating Spectral, Textural, and Morphological Characteristics from UAV Imagery

Le résumé fourni par la source

Above-ground biomass (AGB) is a key parameter to evaluate the growth status of plant population, which is usually expressed by the total mass of vegetation. The accurate measurement of AGB is of great significance in the research and application of precision agriculture. With the advance of sensor and auto-control technology, Unmanned Aerial Vehicle (UAV) equipped with various cameras demonstrates its superior characteristics of high efficiency, low cost, flexible operation, greatly reducing the environmental impact of human measurement and is more suitable for complex farmland environments. Therefore, this paper develops an AGB estimation method of peanut by integrating spectral, textural, and morphological characteristics from UAV imagery. In our approach, the spectral, textural and morphological characteristics of peanut canopy are first calculated from UAV multispectral images. Then, the combination of Pearson product-moment correlation coefficient and recursive feature elimination was used to evaluate correlation between different types of canopy characteristics and peanut AGB. Finally, the Support Vector Machine regression model was constructed to determine the optimal estimation of Peanut AGB. The results suggest that the contribution of plant height was the greatest among different types of canopy characteristics. Furthermore, when using the single type of features, the accuracy is R2=0.776, RMSE=0.074. The integration of spectral, textural, and morphological characteristics effectively improved the accuracy of AGB estimation of peanut, with R2 and RMSE of 0.855 and 0.059 respectively, which would provide a valuable viewpoint for monitoring the growth of peanut.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.