A Plant Disease Detection Using Digital Image Processing With Machine Learning

Manju Mathur, Dr. Manoj Gupta

Plant Disease, Machine Learning Technique, Remote Sensing data.

Agriculture forms an important a role of India’s growth. Approx 50% of India’s population depends on agriculture occupation for his or her livelihood. If plant diseases are not controlled within time it will reduce the productivity to large scale. Plant diseases are generally due to pest, insects, pathogens. The proposed research provides the techniques for regularly monitoring the cultivated area and provides the automated disease detection using remote sensing images. The objective of the proposed system is to fast detection of diseases because it spreads rapidly on the outer layer of the leaves. This system works in two steps; first phase is primary phase in which healthy and diseased training data sets are required. The second phase is monitoring the crop and recognition the disease using machine system. For training data sets, low cost and high accessibility of digital RGB cameras are used. In the experimental result , disease identification performance of CNN architectures are to be shown in terms of accuracy, recall, specificity, training time, storage requirement etc.