Neural Network Classifier Based Method for Leaves Disease Detection with Image Processing Technique

The conventional technique for leaf disease detection involves of calling an expert person who can identify the diseases based on his understanding and that costs too much for an ordinary farmer in an emerging country such as India as stated above. Hence an alternative is mandatory for a country like India where a low cost but technology dependent system is required. To fulfill this, purpose a system is proposed which can identify the diseases on the plants with the help of technology. It will take input image from user which is to be processed. It will preprocess the image and then the green pixels from the image is removed which are nothing but the healthy part of the leaf. The GUI development of this project is done in Matlab. These methods result is shown in GUI. In the future work other part is segmented and the useful segments are selected for further analysis which consist of feature extraction and the statistical analysis of those features. After this the final information about disease of that plant is displayed. The classifier used in this proposed system is Neural Network classifier hence once trained such classifier can provide results in better manner compared to the conventional systems.