Efficient Iris Recognition Preprocessing Algorithm

Iris recognition technology has the advantages of high precision,high anti-counterfeiting and non-invasiveness,and has broad application prospects in many fields such as military,national defense,security inspection and identity verification.Firstly,based on Sobel operator filtering,weighted Hough transform,Gabor wavelet transform,Fourier transform,ellipse fitting,polar coordinate transformation,Rubber Sheet transformation and other methods,an efficient iris segmentation algorithm is proposed to obtain the inner and outer boundaries of the iris.Based on Gabor wavelet transform,hierarchical filtering,directional filtering,dilation and erosion,an eyelid detection algorithm is proposed to obtain the positions of the upper and lower eyelids,which is used to judge whether the available iris region is sufficient for iris feature extraction.Finally,a canthus detection algorithm is proposed based on RCPR algorithm to obtain the position of canthus,which can be used to correct the normalized image under the condition of human eye rotation to improve the similarity of the normalized image of the same human eye.The whole iris pretreatment algorithm designed in this paper can effectively remove the eye diagram that fails the iris recognition due to insufficient image clarity.At the same time,it can reduce the interference of the rotation of the eye of the collected person on the recognition result.Based on the CASIA database and the ZY database collected by myself,this paper tests more than 20,000 images.The experiment shows that the whole algorithm is more efficient and can assist the subsequent iris recognition feature extraction module.