원문정보
초록
영어
In this paper we propose a Hamming Distance Deviation Matching Approach (HDDMA) for Iris recognition. Our HDDMA approach is different from the traditional iris matching method based on Hamming Distance. Firstly we use the odd symmetry Gabor filters with single frequency and two directions to extract iris edge information. Secondly we use zero-crossing detecting method to encode the filtering results． Finally we construct the HDDMA parameter for iris matching. Comparison experiments between the traditional Hamming Distance matching method and the proposed HDDMA are conducted on five iris datasets. The experimental results show that the equal error rate and the correct recognition ratio of the HDDMA are better than those of the traditional Hamming Distance matching method consistently in all iris datasets and the HDDMA has strong anti-eyelid and eyelash noise capability．
목차
1. Introduction
2. Iris Pre-Processing
3. Iris Feature Extraction
3.1. Filter Selection
3.2. Building Filter Set
4. The Zero-Crossing Coding Approach
5. Iris Matching Method
5.1. Traditional Matching Methods
5.2. HDDMA Iris Matching Method
6. Experiments and the Result Analysis
6.1. Experiment Environment and Iris Identify Standard
6.2. Experiment Comparison
6.3. Comparison with other Iris Recognition Methods based on Gabor Filter
7. Conclusion
Acknowledgements
References