Adaptive Iris Recognition Based on the Morlet Wavelet Transform Coefficients
Zhonghua Lin · Guangdian gongcheng · 2009
The theories of the traditional iris recognition methods are complex, and the realizations are difficult. In order to ensure the recognition rate, simplify the algorithm and improve iris recognition efficiency, an adaptive iris recognition method is presented based on the Morlet wavelet transform coefficients. Firstly, locate the iris and fit the contour of lower eyelid, then make normalization to the iris image for getting 512 columns multiplying 64 rows rectangular iris image, and ensure the effective iris area adaptively. Secondly, make one dimension Morlet wavelet transform row by row to the iris image in the effective iris area, get a series of wavelet transform coefficients of different scales and get the distribution figure of these coefficients of different scales. Thirdly, make binary codes to the iris image according to the coefficients of different scales and figure the iris pattern by iris codes. Finally, classify the different iris patterns by adaptive pattern matching method and give the recognition results. Many experiments show that the recognition rates of this method can reach 99.946%, which can meet the demands of iris recognition.