Automating Detection and Management of Cotton Plant Diseases using Deep Learning Methods

When implemented, this project is to design a DL-based module to diagnose diseases in cotton plants and recommended preventive and response measures. Using a diverse dataset spanning multiple cotton varieties, climatic zones, and regions. It will also focus on the gathering and labelling of a different dataset, the preparation of the image dataset, as well as, training of deep neural networks to high levels of accuracy and transferability. In addition, there is an introduction of decision support of the algorithms that provide recommendations depending on the identified diseases. The proposed solution intends to improve the detection of diseases to reduce reliance on manual inspections while offering direction on how to manage a cotton crop and increase yield. With a large dataset that has various cotton types, climate, and areas, the proposed system will use CNNs to detect and categorize various cotton plant diseases from images. It will also focus on the gathering and labelling of a different dataset, the preparation of the image dataset, as well as, training of deep neural networks to high levels of accuracy and transferability. In addition, there is an introduction of decision support of the algorithms that provide recommendations depending on the identified diseases. The proposed solution intends to improve the detection of diseases to reduce reliance on manual inspections while offering direction on how to manage a cotton crop and increase yield.