Corn Leaf Disease Detection using Deep Learning Techniques

The agricultural productivity is the major sector of Indian economy. The rate of misdiagnosis is extremely high, and conventional approaches for identifying crop diseases are ineffective in terms of both time and money. Accurate disease diagnosis also improves economic loss and decreases the effectiveness of agricultural output. The deep neural networks technology have shown promising results in the early phases of machine learning, allowing experimenter for increasing the precision in detection of object and identification systems. The major goal of experiment was to locate and classify maize leaf diseases using appropriate deep-learning-based architectures. This proposed work employ the known dataset of the Corn leaf infection dataset, open source, in trials where this carried work must discriminate between unhealthy and diseased leaves as this exploration where specific areas of the leaf are affected. This proposed work employed block model of VGG for identifying classification problem, kept the most possible number of parameters to be low for achieving rapid convergence while maintaining good accuracy. To locate the contaminated area in the maize leaf, this constructed model applied the cutting-edge deep learning model VGG 16. To show the effectiveness of our suggested solution, this experiment runs on several end-to-end tests and analyses the results. The outcomes demonstrate that the proposed model performs rather existing method given the available data. The proposed model achieves mean average detection precision of 55.30% and a classification accuracy of 95.25%.