Cassava Leaf Diseases Detection and Classification with Advanced Machine Learning Techniques

Cassava leaves offer therapeutic properties and is a great source of food and nourishment. For populations that depend on the crop, however, cassava leaf diseases like Cassava Blight Disease (CBD) and Cassava Mosaic Disease (CMD) can have significant economic and social repercussions. This research tries to identify these diseases in order to prevent losses accurately. The effectiveness of pre-processing techniques like backdrop removal is illustrated as the model is tested with and without background photos. The greatest accuracy for healthy and CMD leaves on the target picture stored in the database with backdrop is 97.32%, while without background, it rises to 99.11% for CMD leaves. For the target image outside the database with background, the maximum accuracy for the CBD leaf is 96.46% and increases to 99.11% for the CBD leaf. Overall, the development of this model could be very effective for the diagnosis and management of cassava diseases, which can have devastating effects on crop yields and food security in affected regions.