Digital Image Segmentation using Rule-Base Classifier

Color image segmentation has become popular in many applications including object classification and recognition, video surveillance, image indexing and retrieval in image databases, and feature based video compression. In this paper novel colour image segmentation is proposed and implemented using fuzzy inference system. This system, which is designed by neuro-adaptive learning technique, apply a sample image as an input and can reveals the likelihood of being a special colour for each pixel through the image. The intensity of each pixel shows this likelihood in the grey level output image. After choosing threshold value, a binary image is obtained, which can be applied as a mask to segment desired colour in input image. To show the efficiency of the proposed method, we use it to segment skin colour, and compare obtained results to some famous algorithms. With 96.8% true positive and 25% false positive, the accuracy of system is proved.

Image segmentation, Colour space, Fuzzy inference system.