Deep Learning-based Nutrient Deficiency symptoms in plant leaves using Digital Images

A novel approach of detecting nutrient deficients of plants based on leaf images has been introduced. One approach involves breaking down the original leaf image into smaller segments and examining each segment using custom-designed convolutional neural networks (CNNs) that have been trained to recognize indications of nutrient deficiencies. The responses from all the CNNs are integrated using a winner-take-all strategy to generate a single response for each block. The final response for the entire leaf is obtained by integrating the responses from all the blocks using a multi-layer perceptron. The method was tested on a sample of 3,000 leaf pictures and was discovered to be more effective than human professionals in detecting nutrient deficiencies. Plants heavily depend on soil nutrients, and when there is a deficiency, some nutrients can be transported from older tissue to newer tissue. To understand the effects of nitrogen deficiency, phosphorus deficiency, and complete nutrient deficiency on plant growth, a study was conducted over four weeks. The findings revealed that all three deficiencies had a significant impact on growth capacity, and there were evident variations in chlorophyll content between the nitrogen- and phosphorus-deficient treatments compared to the complete nutrient treatment. These outcomes suggest that plant survival under environmental deficiencies cannot rely entirely on nutrient mobility and must adapt to thrive.