Multiple crop disease detection using EfficientNetB0 – a CNN model

Abstract Agriculture, the backbone of livelihood as well as the global food security system. Nowadays, it is heavily affected by several diseases. Manually detecting and recognizing diseases in crops in an extensive range is a complicated task, as it requires a great deal of labor and vast experience. To prevent damage to large-scale productions, this project aims to propose the use of Deep Learning approach. A powerful CNN model used to extract features from the dataset provided, and based on that model trained and evaluated. This image segmenting technique used in the model evaluation on the parameters-accuracy, recall, precision, & F1-score and differentiating the disease plants and the healthy ones. The trained model provided the tremendous 99.9 % accuracy in detecting wheat diseases. This approach will help in further advancements as well as help the farmers in safeguarding the crops.