A CNN-Based Deep Learning Method for Accurate Restrain and Classification of Tomato Leaf Disease

Tomato leaf disease detection is an essential task in agriculture to maintain the health and productivity of tomato crops which are a significant component of the global food supply. In the past, diagnosis has involved a form of examination of organs, tissues, or samples by doctors or pathologists which can be a tiresome process and results from which are most of the time prone to high levels of error. Formally, for such tasks, this was performed manually; however, with the development of deep learning technologies, quantization of this process is being performed regularly to increase its efficiency. It is this survey that expects a synthesized picture of the current literature concerning the procedure that seeks to employ deep learning techniques to diagnose diseases affecting tomato leaves. The approaches used in the survey make it possible to achieve high accuracy rates of the deep learning models; some of the models are characterized by an accuracy of more than 95%; thus, can be applied to practical use in the agricultural industry. It also explains some of the issues like lack of supply of data, much complexity of the model and computational resources that are needed. Future research studies should therefore need to take up the issues pointed above and construct even larger databases and likewise try to incorporate the models in real time. Based on the review of literature this qualitative analysis gives an insight about the growing trends and patterns of using deep learning approach for detection of tomato leaf disease and the possibilities towards getting enlightened to further improvement in the agricultural technology in future.