School of Post Graduate Studies Arba Minch Institute of Technology Faculty of Computing and Software Engineering Department of Computer Science Thesis Title Multiclass Support Vector Machine Based Image Processing Model for Enset Diseases Diagnosis

Abstract

Abstract The identification of diseases on the plant is a very important to provide large quantity and good qualitative agricultural products. Enset is an important food crops produced in Ethiopia with great role in food security especially for Southern and South Western parts of the country. There are several issues and diseases which try to decline the yield with quality. Particularly, diagnosis of potential diseases on Enset is based on traditional ways. The visual observation of the expert’s is the main approach that commonly used for diagnosis of such plant diseases. As a matter of facts, visual detections may have defects in terms of accuracy in detection along with lower precision. The aim of this study is to design a model for Enset diseases diagnosis using Multiclass SVM based Image processing techniques. This study presented a general process model to classify a given Enset leaf image as normal or infected. The general structure of the proposed model is including image processing, classification based on Kernel SVM and prediction of recognized disease class. The strategy of K-fold stratified cross validation was used to enhance generalization of Multiclass SVM. Here, the disease diagnosis could apply K-means clustering, color distribution, shape measurements, Gabor texture extraction and wavelet transform as key approaches for image processing technique and Kernel SVM used as a classifier. The researcher chose two common Enset leaf diseases viz. Bacterial Wilt and Fusarium Wilt disease as an infected Enset leaf and collected 430 Enset leaf images (60 healthy and 370 infected) from Areka agricultural research center and some selected fields in SNNPR. For this research work MATLAB version R2017a tool was used as a platform to simulate the real world data. The researcher performed the proposed model based on the analysis of the experimental results with four different kernels, and found that the RBF Kernel achieves the highest classification accuracy as 94.041% and 92.44% for bacterial wilt and fusarium wilt respectively. There is a potential need for IT based solutions to support the manual diagnosis of Enset crop diseases so as to optimize the accuracy for remedial action.