Artificial bee colony optimization (ABC) for grape leaves disease detection

Plants are one of the important sources of food to the mankind. Plant diseases affect the economy of the country and also reduce the quality of agricultural product. Throughout the world, grapes are treated as a major fruit source, which contains nutrients like vitamin C. Diseases in the grape plants seriously affect the quality and production of grapes. Computer vision-based techniques have been used successfully in the last decades, to detect and classify the plants diseases. This paper deals with the identification and classification of diseases in grape leaves by using artificial bee colony (ABC) based feature selection. Initially the input images are applied with pre-processing steps which eliminate noises and background of the images. The features of color, texture and shape are extracted. ABC based attribute selection is used to find the optimal feature set. The selected features are given to the support vector machine classifier for foliar disease detection of grapes. The proposed method is experimented and compared with the other feature selection algorithms. The comparison results depict, the efficacy of the proposed method through the performance metrics such as recall, precision and classification accuracy.