Method of Reducing Dimensions of Segmentation Feature parameter Applied to Skin Erythema Image Segmentation

摘要

In this paper we discussed the method to reduce dimensions of segmentation feature applied to skin erythema image segmentation. Basing on analyzing a lot of erythema images, we proposed a new segmentation feature parameter applied to skin erythema image segmentation. At last, by using the fuzzy-c-means cluster arithmetic, we compared the segmentation speed and segmentation effect of adopting different segmentation features parameter and found that the segmentation result by using the new segmentation feature parameter is better