Krishobondhu : an automated system for diagnosis of paddy disease

Abstract

This paper demonstrates ‘Krishokbondhu’, an automated system for the farmers to identify paddy diseases using their mobile phones. The uploaded pictures captured by the mobile phones are processed in the remote server and presented to an expert group for their opinion. Computer vision techniques are used for detection of affected spots from the image and their classification. A simple color difference based approach is followed for segmentation of the disease affected lesions. Blob detection algorithm is used for feature extraction from the segmented images where features like number of blobs in the crop, nitrogen level of the leaf, area and color values of the affected areas etc are used for classification of the diseases. Binary Search Tree is used for mapping the feature values for comparison of Euclidean distance during classification. The system allows the expert to evaluate the analysis results and provide feedbacks to the famers through a notification to their mobile phones. The mobile application has been developed for Windows Phone and the remote server application is developed using .NET framework.