Palmprint Feature Extraction Based on Improved Gabor Wavelet Combing LBP Operator

摘要

We proposed in this paper an accurate and efficient approach for palm-print feature extraction based on improved Gabor wavelet combing LBP operator. Firstly, the spectrum measured is used to accurately select the direction of Gabor filter, where the amplitude of the Gabor wavelet is directly convolved with the palmprints to obtain Gabor images. Secondly, LBP operator is applied to get the texture images, while the histogram information is extracted as the characteristics of the palmprints. The experiments based on CASIA Palmprint Database show that this method can achieve higher recognition accuracy, more effective and more feasible.