A Systematic Review on Cotton Plant Disease Detection & Classification Using Machine & Deep Learning Approach

Crop diseases, which have a negative impact on food supply as well as economic losses, have a significant impact on crop production and quality yields. In most rural areas of India, agriculture is the main profession. Therefore, it is crucial to use cutting-edge and precise computer vision-based algorithms for automatic crop disease identification and categorization so that prompt prophylactic steps may be advised. Numerous computer vision-based strategies have been reviewed in the literature, along with their benefits and drawbacks. These techniques use a variety of machine learning, deep learning, CNN, and image processing methods. In this study, we thoroughly analyzed recent research studies on the identification and categorization of various cotton plant diseases, conducted by a variety of academics and researchers, and summarized them based on crucial factors, such as the deep learning/machine learning algorithm used, the dataset used for experiments, performance matrices, types of disease detected and classified, and the highest accuracy achieved by the model. As per the analysis carried out, Custom CNN technique along with ResNet 50 / GoogLeNet provides maximum accuracy of 98.53%. In general, the average accuracy of deep learning model is quite high as compare to machine learning model.