A Hybrid Approch Tomato Diseases Detection At Early Stage

Abstract

In traditional farming practice, skilled people are hired to manually examine the land and detect the presence of diseases through visual inspection, but the visual inspection method is ineffective. High accuracy of disease detection is one of the most important factors in crop production and reducing crop losses. Meanwhile, the evolution of deep convolutional neural networks for image classification has rapidly improved the accuracy of object detection, classification and system recognition. Previous tomato detection methods based on faster region convolutional neural network (RCNN) are less efficient in terms of accuracy. Researchers have used many methods to detect tomato leaf diseases, but their accuracy is not optimal. This study presents a Faster RCNN-based deep learning model for the detection of three tomato leaf diseases (late blight, mosaic virus, and leaf septoria). The methodology presented in this paper consists of four main steps. The first step is pre-processing. At the second stage, segmentation was done using fuzzy C Means. In the third step, feature extraction was performed with ResNet 50. In the fourth step, classification was performed with Faster RCNN to detect tomato leaf diseases. Two evaluation parameters precision and accuracy are used to compare the proposed model with other existing approaches. The proposed model has the highest accuracy of 98.6% in detecting tomato leaf diseases. In addition, the work can be extended to train the model for other types of tomato diseases, such as leaf mold, spider mites, as well as to detect diseases of other crops, such as potatoes, peanuts, etc.