Coordinate Attention Mechanism with Convolutional Neural Network for Tomato Leaf Diseases

The tomatoes are worldwide grown important vegetable and taken as economic pillar of several states. Although, plants are susceptible to diversity of weakness which minimize and affect generation of healthy plant, made accurate and initial identification of this disease is challenging. By using of cutting-edge and computer vision gives a solution of this problems. In this research, the Coordinate Attention Mechanism (CAM) with Convolutional Neural Network (CNN) is proposed for detecting tomato leaf disease. The dataset utilized for the research is plant village dataset and data pre-processing are performed. The detection and classification of tomato leaf disease is performed by proposed CAM with CNN method. The proposed method is evaluated with performance measures of accuracy (%), precision (%), recall (%) and f1-score (%). The proposed method attained huge accuracy of 99.98%, precision of 99.45% and recall of 99.53% which is superior than other existing methods like Convolutional Neural Network (CNN), Multinomial Logistic Regression (MLR), CNN based Genetic Algorithm (GA) and Faster Region based Convolutional Neural Network (Faster-RCNN).