Fruit diseases are the most crucial problem in economic losses and production in the agricultural industry worldwide. The apple fruit production industry also has been affected by this problem. This paper proposes an image processing approach for identifying apple fruit diseases. According to the current situation, the farmers take the field officers' disease identification and treatment details. But it takes a few days. So, this proposed system can be used to identify apple fruit diseases quickly and automatically. Healthy and three types of apple fruit diseases, namely,

• apple cod
• flyspeck
• powdery mildew were identified using image processing techniques in this approach. Matlab was used as the main implementing software. This proposed approach is composed of the following main steps. Image Acquisition, Image Preprocessing, Image Segmentation & Feature extraction and finally Classification. Gray-level slicing, RGB channel separation Robert's edge detection, and defected area ratio calculation were used in image segmentation. Before the segmentation, images were preprocessed to a predefined resolution using nearest-neighbor interpolation. Then identify and detect each disease using defected area ratio value. Simple mathematical calculations and mid-level image processing techniques learned in the classroom were used in this approach. Also, we hope to improve accuracy of this application furthermore using machine learning in a very practical manner.

All the implementation details and process workflow of the project can be viewed using this link

This project is still in it's early stages. We hope to implement Machine Learning algorithms to further improve this project. Because currently values used to classify between above mentioned diseases are manually obtained by us and not using Machine Learning. Our dataset is also very small.