Research on Application of Iris Recognition Based on Improved Algorithm
Miaomiao Huang · Computer Technology and Development · 2013
The appropriate threshold is obtained through the histogram of the iris image,which is used in iris binary processing.The iris inner boundary is orientated roughly based on geometric gray-level projection method and the iris external boundary is orientated by the improved Daugman localization operator,both of which are fitted accurately by the method of the least squares round fitting.The experimental results show that the localization algorithm has the advantages of high positioning accuracy and high positioning speed,and can satisfy the requirement of real-time.Meanwhile,put forward an iris normalization method which is based on line extraction.In the feature extraction,iris feature information is extracted by the method of 2D Log-Gabo filter which can generate multi-channel filter though multi-scale and multiple directions,the method also can reduce the length of the iris code and the time of feature matching,improve the feature matching accuracy,can effectively improve the accuracy of iris recognition.