利用一維特徵對不同角度、表情與光源之臉部影像進行辨識

Face Recognition: Using 1-Dimension Feature

林士譽(Lin Shih-Yu)

指導教授 ： 李忠謀教授

國立臺灣師範大學/教育學院/資訊教育學系/碩士(2001年)

摘要

本研究提出一個可不受不同拍攝角度、表情或光源影響之臉部影像辨識方法。我們對臉部影像進行兩次影像二元化後，再將臉部輪廓橢圓化以擷取臉部輪廓。而利用臉部特徵之位置關係，即可定位出理想且具有不受外物遮蔽及包含重要臉部資訊等特性之臉部中心點─鼻尖。而臉部的特徵取自於從臉部中心點至臉部輪廓之lattices，人臉之辨識則簡化為lattice特徵之比對。本實驗以ORL臉部影像資料庫內之影像作為實驗資料，實驗結果顯示，上述理論方法能有98%之正確辨識率，而在加入自動取得臉部中心點與臉部輪廓計算過程中之誤差後，辨識率仍能維持在90~93%之間。

關鍵字

臉部辨識 ； 一維特徵

並列摘要

This thesis proposes a face recognition method under varying pose, facial expression and lighting conditions. Binary thresholding techniques were used to identify important facial regions before fitting of ellipsoid to extract facial boundary. Using relative positional information of eyes and nose, the nose region is assumed and the nose top is localized and defined to be the center of the face. Lattice features between the face center and the boundary of the face are then computed. Face recognition is achieved by lattice matching. For the experiment, the ORL face database is used. Our experiments showed that 98% recognition rate could be achieved with the proposed face recognition model with precise feature extractions (human interactions). However, given our method for facial center localization and facial boundary detection, recognition rate of 93% can still be achieved.

並列關鍵字

face recognition ； one-dimensional features

參考文獻

[29] Y. S. Huang, Y. H. Tsai, C. M. PengWu, C. Y. Liu, S. W. Jeng, C. C. Chang, “A novel light compensation approach based on subtracting background illumination intensity distribution for face recognition,” Proc of 2001 CVGIP, 2001.

[1] A. S. Georghiades, P. N. Belhumeur, and D. J. Kriegman, “ From few to many: illumination cone models for face recognition under variable lighting and pose,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 23, pp. 643-660, 2001.

[2] A. Shashua, and T. Riklin-Raviv, “ The quotient image: class-based re-rendering and recognition with varying illuminations,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 23, pp. 129-139, 2001.

[3] A. S. Tolba and A. N. Abu-Rezq, “Combined classifiers for invariant face recognition,” Proc. of International Conference on 1999, pp. 350-359, 1999.

[4] C. Liu, H. Wechsler, ”A shape- and texture- based enhanced fisher classifier for face recognition,” IEEE Trans. on Image Processing, vol. 10, pp. 598-608, 2001.