Optical nerve disc segmentation using circual integro differencial operator

Abstract

The paper deals with design of suitable methodology for the detection of the optic nerve disc from ophthalmological images. This issue was frequently solved in the past based on the Hough transform. The problem is that some the clinical images lack of sufficient contrast and so it is difficult to identify individual objects in the images. In our analysis, we process the image data, which are poorly distinguishable. Another significant drawback is the lower resolution of the analyzed data, which leads to poor identification of the optic nerve disc. For the detection of the optic nerve disc is used integro differential operator, which is looking for objects in the image that have a circular shape. The advantage of this proposed solution is sufficient sensitivity even in the weaker contrast environment.