Continuous advancement in image processing and machine learning techniques have made itrnpossible for computers to see and learn. What is seen by the eyes of human beings could be dividedrninto pixels and given to a computer so that the computer will be able to see and learn based on thernprovided values. Based on the input values fed in computers could learn to identify various thingsrnbased on the things they have been learnt from them. There are many possible areas in whichrncomputers can be applied to see and learn in order to make the life of human beings much easier.rnIn this thesis an approach has been proposed which is capable of automatically detecting andrnclassifying plant disease from an image based on artificial neural network. Now a days, plants havernbecome much more important than they used to some years ago where they have been only usedrnto feed mankind as well as animals. Plant diseases are currently detected and classified usingrnmethods that requires a lot of manual work with experts, agricultural extension worker and farmersrnwhich is both time consuming and error prone. To automate the process of plant disease detectionrnand classification different researchers have studied many techniques using both machine learningrnand image processing. However, these proposed techniques still have limitation.rnThe steps followed in this research for detecting and classifying the plant disease are: datasetrncollection, image pre-processing, masking, and removing the green part, feature extraction andrnselection, classification, and disease management techniques. For comparing and demonstratingrnthe conventional machine learning techniques and proposed approach respectively two differentrntypes of plant have been selected namely, maize, and potato from the plantvillage.org website.rnSince the conventional machine learning techniques do not have the potential to extract and selectrnfeatures from a given raw data, texture features using Haralickâ€™s from color co-occurrence matrixrnhave been extracted and selected using subset feature selection technique.rnThe proposed approach and the selected conventional machine learning techniques were evaluatedrnusing confusion matrix, classification performance report, and t-test to asses which has the higherrnclassification potential. The proposed approach achieved an average accuracy of 97.6%, averagernprecision of 97.0%, 97.0% of average recall ,and average F1 value of 97.0% over a test dataset ofrnpreviously unseen 1201 images. From the analysis of the experimental results the proposedrnapproach gives best result than the conventional machine learning classifier. This due to the factrnthat convolutional neural network extract high level features from the input raw data, making itrnmore efficient and accurate, and avoid errors due to a subjective manual feature extraction therebyrnshowing the feasibility of its usage in real time applications for the classification of healthy andrnnon-healthy plants.