Leveraging Image Analysis for High-throughput Phenotyping of Legume Plants

Background: The advancements achieved in artificial intelligence (AI) technology in recent decades have not yet been equaled by agricultural phenotyping approaches that are both rapid and precise. Efficient crop phenotyping technologies are necessary to enhance crop improvement endeavors in order to fulfill the projected demand for food in future. Methods: This work demonstrates a method for non-destructive physiological state phenotyping of plants using cutting-edge image processing methods in conjunction with chlorophyll fluorescence imaging. Key fluorescence metrics, such as fv/fm and NPQ, were extracted from images taken at different phases of development via processing. In addition, this research explores the transformative role of automated image analysis in high-throughput phenotyping of legume traits. A comprehensive examination of recent studies reveals the diverse applications of machine learning and deep learning algorithms in capturing morphological traits, assessing physiological parameters, detecting stress and diseases in various legume species. The comparative analysis underscores the superiority of automated systems over traditional methods, emphasizing scalability and efficiency. Challenges, including algorithm sensitivity and environmental variability, are identified, urging further refinement. Recommendations advocate for standardized metrics, interdisciplinary collaborations and user-friendly platforms to enhance accessibility. As the field evolves, the integration of automated image analysis holds promise for revolutionizing legume phenotyping, accelerating crop improvement and contributing to global food security in sustainable agriculture. Result: The findings demonstrate that the proposed method is effective in illuminating how plants respond to their environment, hence promoting advancements in plant phenotyping and agricultural research.