Abstract: The very initial phase to alleviating losses in the quality and productivity of agricultural products is recognizing plant diseases. Utilizing these modern innovations is a key objective of this effort, which focuses on plant disease detection specifically. The system will use state-of-the-art image processing algorithms to evaluate plant pictures in order to accurately detect diseases and provide suggestions. for the necessary nutrient supplements and pesticides. The suggested technique will not only detect the presence of diseases but will additionally pinpoint specific nutrient deficiencies that have caused issues with crop health. In the end, this will support prompt and effective management of crop diseases, increase agricultural productivity, and lessen financial losses. Equipped with 13505 photos of crop leaves from an available dataset, a Residual Network (ResNet-9) was trained to deal with this classification operation. The postulated ResNet-9 model showcased its viability by achieving 99.20% accuracy on a test set. All in all, employing an open-picture dataset to train ResNet models proposes a reliable path toward crop disease detection.

| DOI: 10.17148/IJIREEICE.2024.12609