Revitalizing Paddy Yields with Computer Vision

A groundbreaking innovation in the field of plant science revolutionizes the precise identification of nutrient deficiencies in paddy plants through meticulous leaf analysis. The methodology commences by breaking down input leaf images into manageable blocks, subsequently subjecting them to rigorous evaluation using specialized convolutional neural networks (CNNs), each meticulously trained to discern specific nutrient deficiencies. The advanced network architecture systematically scrutinizes these blocks for characteristic deficiency symptoms. To achieve an all-encompassing evaluation of the leaf's condition, the responses generated by all CNNs are aggregated, and a multi-layer perceptron synthesizes them into a singular block response. The proposed approach results in a comprehensive assessment of the overall leaf health. In a comprehensive study encompassing seven different deficiency types - including Nitrogen (N), Phosphorus (P), Potassium (K), and Zinc (Zn) deficiencies - a vast dataset comprising 2,750 carefully collected leaf images was employed. The results demonstrate the proposed method's superiority, surpassing the accuracy of expert human assessments in identifying various nutrient deficiencies. The proposed method harnesses the power of CNNs, effectively merging the image analysis, machine learning, and Convolutional Neural Networks (CNNs) to propel nutrient deficiency analysis in paddy plants to unprecedented heights.