| Summary: | 碩士 === 義守大學 === 資訊工程學系碩士班 === 96 === Biometric identification means using measurable physiological features to identify or verify a user’s identity. There are many applications of biometric identification, e.g. security control, verification of the staff identities in companies. At present, there are several methods for biometric identification, including fingerprint, iris, face recognition, and so on. Although these methods possess high recognition rates, they still encounter the problems of high cost, high complexity of features, and personal privacy. In this thesis, we propose an adaptive biometric system based on the palm features. A palm image is acquired by a scanner. Then, six local sub-images of palm are found out using palm feature points and a correction procedure. After that, the feature vectors of these sub-images are extracted by a method of local fuzzy patterns and used to train a learning vector quantization neural network. Finally, an unknown user can be identified or verified by the neural network. Experimental results have shown that our method can extracted the palm feature points correctly and possesses a well description of texture features and a leaning capability.
|
|---|
