Water status and plant traits of dry bean assessment using integrated spectral reflectance and RGB image indices with artificial intelligence.

Abstract

This study investigated the potential of using remote sensing indices with artificial neural networks (ANNs) to quantify the responses of dry bean plants to water stress. Two field experiments were conducted with three irrigation regimes: 100% (B100), 75% (B75), and 50% (B50) of the full irrigation requirements. Various measured parameters including, wet biomass (WB), dry biomass (DB), canopy moisture content (CMC), soil plant analysis development (SPAD), and soil water content (SWC) as well as seed yield (SY) were evaluated. The results showed that the highest values for WB, DB, CMC, SWC, and SY were achieved under B100, while the highest SPAD values were achieved under B75. The study also found that most of the RGB image indices (RGBIs) and spectral reflectance indices (SRIs) exhibited a linear relationship with the measured parameters and SY, with R² values ranging from 0.34 to 0.95. In contrast, SPAD showed a significant quadratic relationship, with R² values ranging from 0.34 to 0.79. Additionality, the newly developed SRIs demonstrated 5-40% higher correlations compared to the best-performing published SRIs across all measured parameters and SY. ANNs using RGBIs and SRIs separately demonstrated high prediction accuracy with R 2 values ranging from 0.79 to 0.97 and 0.86 to 0.97, respectively. Combining the RGBIs and SRIs, the ANNs achieved higher prediction accuracy, with R² values ranging from 0.88 to 0.99 across different parameters. In conclusion, this study demonstrates the effectiveness of using SRIs and RGBIs with ANNs as practical tools for managing the growth and production of dry bean crops under deficit irrigation.

Plant phenotyping relevance

RGB画像指標・スペクトル反射指標とANNを用いて、乾燥豆のバイオマス、水分、SPAD、収量などの植物形質を定量・予測する手法を開発し、相関および予測精度を評価している。形質取得・推定法が研究の中心である。

Code and data availability

The paper's phenotype dataset (WB, DB, CMC, SPAD, SWC, SY with SRIs and RGBIs) is not publicly deposited; the authors state it is available only upon request. No public code, images, or model checkpoints are mentioned. The NASA POWER Data Access Viewer is an external meteorological data source, not a paper-specific pha

No evidence-backed public reproduction asset is currently recorded.