AI Models for Advancing Plant Phenotyping and Phenomics

Plant phenomics has advanced rapidly with the emergence of high-throughput phenotyping (HTP) technologies, which enable the non-destructive measurement of morphological, physiological, and biochemical traits across diverse growth stages and environments on a large scale. These platforms generate complex, multimodal datasets, shifting the primary bottleneck from data acquisition to efficient and biologically meaningful data analysis. In this context, artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), has become a cornerstone of modern phenomics by enabling automated, accurate, and scalable trait extraction. AI-based approaches have demonstrated strong performance in phenotyping applications spanning plant development, shoot and root architecture, growth dynamics, stress responses, and yield-related trait estimation at organ, whole-plant, and canopy scales. The integration of AI with non-invasive imaging technologies has further accelerated phenomics by supporting the development of open-source tools, data management frameworks, and large, community-driven datasets. Despite these advances, significant challenges remain, including high platform costs, limited field adaptability, scarce annotated datasets, domain shifts between controlled and field conditions, and limited model interpretability. This chapter reviews recent advances in plant phenomics enabled by HTP and AI, with emphasis on developmental, root, shoot, and yield phenotyping, as well as applications in biotic and abiotic stress assessment. Current limitations and emerging strategies for cost-effective, field-deployable, and interpretable AI integration are also discussed, underscoring the role of AI as a critical driver of scalable phenotyping and accelerated crop improvement.