An Approach for Eggplant Disease Recognition

Abstract

Bangladesh's agricultural sector employs the most people. Because Bangladesh is an agricultural country, agricultural goods are the main source of income for the majority of Bangladeshis. The biggest barriers to the growth of our country's agricultural industry are a lack of opportunities and infrastructure, natural calamities, and crop diseases. For farmers of broadacre crops, plant diseases constitute serious productivity and quality limitation. Because significant crops are threatened by a variety of plant diseases and pests, the disease may have an impact on the crop's quality. As a result, it's critical for the farmer to discover the infection at the right time. Crop disease can be effectively monitored and controlled for agricultural and food safety when it can be simply traced out. Because eggplant is a muchneeded plant in our nation, we focused our effort on disease detection. If the eggplant is harmed, the country's economy will be severely harmed. Because eggplant has so many properties, it will have a large negative impact on nutrition if it is transited. To identify crop illness quickly and correctly, a technology called computer vision and deep learning is utilized, and we used it to detect eggplant infections. This technique makes it very lucrative and simple to identify plant diseases. This is due to the fact that it decreases the massive workload of crop monitoring and may identify disease symptoms at an early stage. We employed transfer learning models and were able to reach an average accuracy of 99.06% from DenseNet201, followed by Xception’s average accuracy of 99.04% and ResNet152V2’s average accuracy of 98.93%. We'll achieve this aim by keeping the public in mind and assisting farmers with efficient crop cultivation.