Noise detection of iris segmentation
Xiaoliang Liu · Jisuanji gongcheng yu sheji · 2009
In view of iris noise of the eyelid, eyelash and reflection spot affect performance of iris recognition, a method of noise detection about iris segmentation is proposed. Based on Mathematical morphology, the proposed method uses the least squares method combined with edge detection to determine eyelid occlusion regions, then, eyelash and reflection spot is detected with threshold method. The experiment indicated that this method may detect and suppress these noises accurately and effectively, it is an effective and feasible noise detection algorithm of iris image.