Hybrid Classification Approach with Textural Feature Analysis for Plant Disease Detection

The significant task of image processing in the domain of plant disease is of diagnosing the diseases occurred on plants because much complicated data is employed as input. Different stages are executed to detect the plant diseases using several techniques. The earlier work suggests SVM technique for diagnosing plant infections. This work projects a voting mechanism to improve various parameters like accuracy, precision and recall in comparison with the traditional methods. The projected mechanism was simulated using MATLAB. An analysis is conducted on the results with regard to some metrics. The outcomes exhibited that the projected mechanism is more effective as compared to other method. The proposed model achieves accuracy of 96 percent which approx. 3 percent higher as compared to existing SVM classifier.