Optical EngineeringRetinal blood vessel detection using wavelet-matched filter
|Format||Member Price||Non-Member Price|
A novel method for retinal blood vessel detection by using a wavelet-matched filter is proposed to improve the detection performance. Vessel references are generated by using dual Gaussian functions with different widths, separations, and orientations, while a Mexican hat wavelet is used for feature enhancements. The results show that the proposed method has advantages over the conventional matched filter in that the vessels can be easily distinguished from the fluctuating background without the need of a thresholding process, and vessel images can be precisely reconstructed.