COTTON LEAF DISEASE DETECTION USING ARTIFICIAL NEURAL NETWORK

It is difficult for farmers to observe disease in the naked eye. Different types of plant disease reduce the plant life cycle. The presence of disease in plants can be observed by the symptoms present in the leaf, stem or infection occur in the fruit. Our proposed system gives an effective way for the detection of disease in the plant through the infection occurs in the leaf. The first step in disease detection is capturing cotton leaf disease through the device. In the second step, cotton leaf images are pre-processed through image segmentation. in the third step, RGB components i.e., red, green, and blue are removed from the image of the leaf, then it is converted to an HSV image. In the fourth step, white and black images of the cotton leaf are made; defective parts in the images are indicated by white colour. In the next step, an artificial neural network is trained to distinguish the disease and healthy leaf. By using our proposed system disease can be detected at the early stage with an accuracy of 85%. Disease or infection should be identified at the initial stage otherwise it leads to lossof crops.