Automatic Diabetic Retinopathy detection using OMFR with minimum cross entropy threshold

Retinal blood vessels damaged from diabetic retinopathy can cause vision loss. Diabetic Retinopathy (DR) detection, poor quality retinal image makes more difficult the analysis for ophthalmologist. In this paper presents a new automatic method of blood vessels extraction with a real time process using optimized matched filter with minimum cross entropy threshold. Minimum cross entropy threshold is effective in the threshold technique; it could be very time-consuming in normal or abnormal retinal image for vessels segmentation. The performance of this proposed method was calculated on DRIVE database. This method computational time of whole MATLAB process was taken in a real time for each retinal image.