Exploring reactive oxygen species dynamics in pea germination to improve variety selection in stressful environments

Non-biological stressors constitute a significant threat to crop productivity and food security, underscoring the imperative to select appropriate seeds for specific environmental conditions. However, conventional method for seed variety selection is time-consuming and labour-intensive. This study introduces an AIEgen (TBPBB) based image recognition system, which utilizes fluorescence probe imaging signals of reactive oxygen species (ROS) during the germination phase of pea (Pisum sativum L) seeds to to accelerate the selection of superior varieties under non-biological stressors. The accuracy of the system's screening is highly congruent with that of traditional manual selection methods, with an accuracy of 100% for harvest parameters, 93.8% for germination percentage and 87.5% for vigour index in distinguishing pea varieties in an alkaline cultivation environment. This seed selection strategy significantly reduces the labor intensity, achieving a substantial reduction in time from 228 days to 73 hours. The research underscores the potential of AIEgen-based fluorescence image recognition technology in enhancing crop quality under specific stressors conditions. It offers a promising avenue for innovative technological approaches that contribute to the sustainable development of agriculture.