Detection of Banana Plant Diseases Using Convolutional Neural Network

For ensuring global food security and sustainable agriculture, a major challenge is to control plant diseases. There is a need to improve existing procedures for early detection of plant diseases by using deep learning-based modern automatic image recognition systems. The paper describes such methods based on convolutional neural networks. Crop monitoring is the monitoring of crop growth and performance during developmental stages, and it allows farmers to intervene at the right time to ensure optimal yields at the end of the season. All over the world, banana production is affected by numerous diseases. Most diseases include Panama wilt, leaf spot diseases, yellow Sigatoka, black Sigatoka, bacterial wilt, bunchy top, banana bract mosaic virus, and CMV (Cucumber Mosaic Virus). Innovative and quick methods for detecting diseases will allow us to monitor more efficiently and create proper fertigation strategies, which will help farmers increase their yield. Combining aerial image data from unmanned aerial vehicles (UAVs) with machine learning algorithms can deliver an accurate and efficient technique for detecting crop diseases in real-world conditions. The paper describes one approach for detecting banana plant diseases in the Jalgaon District of India using various deep-learning approaches. Accuracy is achieved by more than 98%.