Classification Method for Macular Lesions Using Fuzzy Thresholding Method

The paper deals with analysis of image data macular lesions. In clinical practice, there is a problem with identification of macular lesion. Those pathological changes are fairly sufficiently observable. On the other hand, there are complications with computing and investigating their area. To the best of our knowledge, there is no any proprietary software for automatic extraction and classification stage of macular lesions. The proposed algorithm offers suitable way for automatic extraction. The core of the algorithm is based on the classification individual pixels to output classes. Each class represents specific object in the image. After taking classification process, individual objects are separated. By this approach we obtain mathematical model of analyzed lesions. Furthermore, segmentation method uses color mapping of shade level images. It means that we obtain color intensity image instead of shade level data.