Scorpion image segmentation system

摘要: Death as a result of scorpion sting has been major public health problem in developing countries. Despite the high rate death sting, little report exists literature intelligent device and system for automatic detection scorpion. This paper proposed digital image processing approach based on floresencing characteristics Scorpion under Ultra-violet (UV) light identification The acquired UV-based images undergo pre-processing to equalize uneven illumination colour space channel separation. extracted channels are then segmented into two non-overlapping classes. It observed that simple thresholding green RGB is sufficient segmenting from other background components image. Two approaches segmentation have also this work, namely, average technique K-means segmentation. algorithm tested over 40 UV obtained different part world results show an accuracy 97.7% correctly classifying pixel clusters. 1system will eliminate associated with some existing manual presently use detection.