ИСТИНА

In this paper, we propose a novel image processing and analysis method for immunofluorescence microscopy images of skin tissue for the problem of pemphigus diagnosis. The method includes illumination equalization algorithm, ridge detection, Euclidean distance transform, graph cut algorithm and cell boundary density metric. First four steps perform image enhancement and with last step we get numerical result. The proposed method helps the doctor find the intercellular structures both visually and numerically and make diagnosis. Experimental results for various patients show conformity between doctor's diagnosis and results of the proposed method.