Convolution Neural Network based Several Orange Leave Disease Detection and Identification Methods: A Review

Agriculture plays an essential role in ecological, social, biological prospects of about 65 per cent to 70 per cent of the human population. The production of the crop decreases due to the bad effects of leaves, fruit leaves and plant diseases. Fruit, leaf infection is the main reason of the maximum damage of the fruit which decrease the amount and superiority of the farming. Citrus canker infection may infect plants or fruits like as orange leaves that disturb the creation of fruit in plants. The early detection of the disease is challenging aspect of improving the productivity of the plant. This analysis emphasized on the identification, classification and detection of the leaf disease. Firstly, describes the concept of the orange, orange leaf disease and symptoms. Then, a detailed survey of various papers has done. In addition, a detection and identification techniques are explained which includes GLCM, multi-SVM, threshold value, neural network and Fuzzy method. Then, a method is explained to diagnose infected orange leaf on the basis of the convolution neural network (CNN). Convolutional neural network (CNN) is one of the main area of research in classification and detection of the leave disease. The features of the fruit (orange) leaves are extracted to detect and classify the disease. In this study, various stages for identification of the leaves disease are described. In addition, different phases to overcome the leaves diseases are also explained namely conjuring up hidden layer method, visualization of feature, semantic library and consideration mapping. Lastly, architecture is demonstrated along with different steps and procedure in orange leaf disease.