Mustard and Mung Bean Diseases and Pest Detection and Classification Using Deep Learning Techniques
摘要
Global crop illness and pests pose a severe danger to global food security. One of the most important crops in India is the mustard plant (Brassica juncea) and the mung bean (Vigna radiata). Therefore, taking care of this type of plant is essential. Like humans, plants are susceptible to several bacterial, fungal, and viral diseases. As a result, a more accurate way of diagnosing plant diseases and pests using deep learning is recommended for agricultural improvement. The farmers can discover plant diseases and pests using the substantial assistance of this research. With this proposed system, we aim to identify pictures using CNN. The classification portion identified the plant’s disease and pest by creating a training database using the original leaf picture. One issue with our system is that few cropped photos of the mong bean and mustard plant are available for training. To overcome this, we used a variety of transfer learning models to obtain good accuracy. This model’s development has been successful in recognizing the five distinct mung bean illnesses and the five different mung bean pests. Additionally, it correctly identifies four disease types and four pest types that affect healthy mustard. Based on the experiment, our deep learning model has a good accuracy of 99% for mustard and 94% for mung bean when it comes to identifying pests and diseases in these two crops.