Fundus lesion identification based on the improved FCM and mathematical morphology

Fundus lesion identification is a worldwide problem. The traditional method has greater impact on human factors, subjective and cumbersome makes Fundus lesion identification Accuracy, objectivity and practicality not be guaranteed. To solve the above problem, an improved FCM algorithm for segmentation of Fundus lesion, the improved FCM algorithm clustering image segmentation Fundus lesion, and then remove the noise by mathematical morphology operations. Experimental results show that the algorithm can effectively identify the lesion in the Fundus image.