
Rice, vital for global food security, faces production challenges during the heading-flowering stage. Traditional phenotyping struggles for large-scale analysis, prompting a shift to advanced computer vision and deep learning. While methods like SIFT and neural networks enhance rice panicle analysis, capturing dynamic growth necessitates merging field cameras with deep learning for precise, real-time monitoring.