Applay Local Gabor Ternary Pattern for video-based Illumination Variable Face Recognition
The illumination variations problem is one of the well-known problems in face recognition in uncontrolled environment. In this paper, we propose a novel approach which combines Gabor filters with LTP operator to address this problem. Gabor feature is robust to illumination variations and has been successfully used in face recognition. Local Ternary Patterns (LTP), a generalization of the Local Binary Pattern(LBP) local texture descriptor that is resistant to illumination effects because it is invariant to monotonic gray-level transformations, and it has been shown to have high discriminative power. In our algorithm, we convolve the image with Gabor filters to extract their corresponding Gabor feature maps firstly, and then use the LTP operates on each Gabor feature maps to extract the local neighbor pattern. Finally, we describe the input face image by using the histogram sequence extracted from all these region patterns. The results compared with the published results on Yale-B and CMU PIE face database of changing illumination verify the validity of the proposed method.
LTP Face Recognition Illumination Gabor Wavelets LGTP video-based
Yong Han Fagen Tang Huafeng Wang
School of Computer Science and Engineering Beihang university Beijing, China
国际会议
2010 International Conference on Software and Computing Technology(2010年软件与计算机技术国际会议 ICSCT 2010)
昆明
英文
193-196
2010-10-17（万方平台首次上网日期，不代表论文的发表时间）