基于二維Gabor小波的人臉識(shí)別算法
Face Recognition Based on Two-Dimensional Gabor Wavelets
-
摘要: 該文提出了一種基于二維Gabor小波的人臉識(shí)別算法。該算法先對(duì)人臉圖像進(jìn)行多分辨率的Gabor小波變換，然后在圖像上放置一組網(wǎng)格結(jié)點(diǎn)，每個(gè)結(jié)點(diǎn)用該結(jié)點(diǎn)處的多尺度Gabor幅度特征描述，采用主元分析法對(duì)每個(gè)結(jié)點(diǎn)進(jìn)行去相關(guān)、降維，最后形成特征結(jié)。把每個(gè)特征結(jié)作為觀測(cè)向量，對(duì)隱馬爾可夫模型進(jìn)行訓(xùn)練，并把優(yōu)化的模型參數(shù)用于人臉識(shí)別。實(shí)驗(yàn)結(jié)果表明，該方法識(shí)別率高，復(fù)雜度較低。
-
關(guān)鍵詞:
- 人臉識(shí)別;Gabor小波變換;主元分析;隱馬爾可夫模型
Abstract: A new approach based on two-dimensional Gabor wavelets transform for face recognition is presented. The Gabor wavelet representation of an image is the convolution of the image with a family of Gabor kernels. A set of vectors called nodes, over a dense grid of image points are formed, and each node is labeled with a set of complex Gabor wavelets coefficients. The magnitudes of the coefficients are used for recognition. Principal component analysis is a decorrelation technique and its primary goal is to project the high dimensional vectors into a lower dimensional space. Feature nodes, as observation vectors of HMM, is derived by using principal component analysis. A set of images representing different instances of the same person is used to train each HMM, and each individual in the database is represented by an optimal HMM face model. Experimental results show that the proposed algorithm has a high recognition rate with relatively low complexity. -
Gong S, Psarrou A. Dynamic Vision: from Images to Face Recognition[M]. London: Imperial College Press, 2000: 5.20.[2]Daugman J. Two-dimensional spectral analysis of cortical receptive field profiles[J]. Vision Research, 1980, 20(10): 847856. .[3]Lades M, Vorbruggen J C, Buhmann J. Distortion invariant object recognition in the dynamic link architecture[J]. IEEETrans. on computers, 1993, 42(3): 300311. .[4]Rabiner L. A tutorial on hidden Markov models and selected application in speech recognition[J].Proce. IEEE.1989,77(2):257-[5]Nefian A. A hidden Markov model-based approach for face detection and recognition[D/D]. Georgia: Georgia Institute of Technology, 1999: 38.108.[6]Othman H, Aboulnasr T. A separable low complexity 2D HMM with application to face recognition[J].IEEE Trans. on Pattern Analysis and Machine Intelligence.2003, 25(10):1229-[7]Helmuth L. Objection recognition: where the brain tells a face from a place[J]. Science, 2001, 292(5515): 196198. .[8]Duda R, Hart P, Stork D. Pattern Classification, second edition[M]. New York: Wiley-Interscience, 2000: 114.139.[9]Samaria F. Face recognition using hidden Markov model[D/D]. Cambridge: University of Cambridge, 1994: 2782. -
計(jì)量
- 文章訪問(wèn)數(shù): 2729
- HTML全文瀏覽量: 123
- PDF下載量: 3335
- 被引次數(shù): 0
下載:
