Deep Learning based Disease Detection in Tomatoes
Sheril Angel. J, Eugine Mary. J, U Dikshna., Blessy Athisaya Malar, A. Diana Andrushia, T. Mary Neebha
Abstract
Agriculture is the common for all creatures in the world. The agriculture products play a vital role in the food commodities. This paper presents the defect detection of tomatoes using deep learning techniques. The RGB tomato images are taken for the experiment. The pre-processing steps of background removal is used to separate the tomatoes from the background. Heuristic threshold based background removal is adhered. The convolution neural network based deep learning method is adopted to detect and classify the tomatoes. Three layers of CNN has used for the disease detection of tomatoes. The real time tomato images are used for this study and achieved 0.971 detection accuracy.