UAV‐based spectral–textural integration improves functional assessment of soil amendment effects in reclaimed agricultural systems

This study evaluates the effectiveness of soil amendment materials and their combinations in reclaimed agricultural systems using unmanned aerial vehicle (UAV)‐derived functional indicators. A field experiment was conducted in a subtropical red soil region, where organic fertilizer, biochar, and oyster shell powder were applied at varying rates and combinations. Rice ( Oryza sativa L.) leaf area index (LAI) and above‐ground biomass (AGB) were measured at key growth stages and modeled using UAV multispectral data integrating spectral information and texture information. The results showed that spectral–textural feature integration improved estimation accuracy compared with spectral features or texture features. The best‐performing AGB models achieved R 2 values of 0.87 with root mean square error of 166.64 g/m 2 and showed high reliability. UAV‐derived AGB effectively captured treatment‐induced differences and provided consistent assessment of soil amendment performance. These findings indicate that UAV‐based spectral–textural integration offers a robust approach for functional evaluation of soil restoration in heterogeneous agricultural systems, with AGB serving as a more reliable indicator than LAI for monitoring soil improvement outcomes.