Automatic detection and measurement of foveal avascular zone

This paper describes a new digital image processing algorithm which is concerned with the automatic detection and measurement of the foveal avascular zone (FAZ) in Fundus Fluorescein Angiographies (FFA). FFA is used in ophthalmic practice to evaluate vascular retimpathies and choroidopathies. In the proposed algorithm, projection in the horizontal and vertical directions will be applied to determine the location of the dark regions in the test image that represents the region of interest. Then, image is segmented based on its gray level values by using thresholding techniques. Morphological operators and Sobel edge detector are then used to detect and localize the contour of the FAZ.