An Efficient Citrus Canker Detection Method based on Contrast Limited Adaptive Histogram Equalization Enhancement

摘要

Plant leaf disease is one of the critical causes for heavy loss of yield that reduces quantity and reduces quality of the agricultural product. Citrus plants such as lemon are mainly affected by citrus canker disease which affects the fruit production of the plants. Early canker disease identification is one of the challenging solution for increasing the plant production. This paper aims to identify and classify the canker disease accurately from the affected leaf images by adopting image processing techniques to detect plant leaf diseases from digital images. The proposed approach involves two stages to improve the clarity of leaf images. The primary stage uses Contrast Limited Adaptive Histogram Equalization (CLAHE) in pre-processing step which improves the contrast level of disease affected leaf image, segment the region of interest using K-mean Clustering and texture feature extraction using statistical GLCM. The second stage by adopting the Support Vector Machine classifier to detect the canker leaf image and implements these methods in lemon citrus canker disease detection. Experimental results show effective accuracy detection and reduced execution time of canker disease detection.