Tomato leaf disease detection using Resnet-50 and MobileNet Architecture

Abstract

Diseases in Tomato mostly on the leaves affect the reduction of both the standard and quantity of agricultural products. Several diseases such as bacterial spot, early blight, late blight, leaf mold, septoria leaf spot, spider mites, two-spotted spider mite, yellow leaf curl Virus, mosaic virus common diseases found in tomato, thus, real-time and precise recognition technology is essential. To detect plant leaf diseases, image processing techniques such as image acquisition, segmentation through two technical models Resent50 and MobileNet are implemented. These two methods are implemented by the transfer learning method which widely used for deep learning, where every step is get improved than the previous one. The deeper stages the execution goes, the more accurate result tends to yield. In Resnet-50 Model, experimental results fluctuate from 94 percent to 99.81% and In MobileNet the predictions correction resonates within 95.23% to a maximum of 99.88% which buttress the prediction with respect to the actual data by analyzing accuracy and execution time to identify leaf diseases.