應用二維主成分分析之人臉辨識

人臉辨識(Face Recognition)在模式識別領域中一直是個很重要的主題，它將影像、圖片或是攝影機中偵測出來的人物，識別出其身分。應用範圍包含「數位監控系統」、「門禁管理」、「智慧型人機互動」、「犯罪偵查」、「出入口管制」、「個人化的服務系統」等。基於所取得的人臉影像在傾斜角度、場景光線、髮型或是表情都可能有不同的呈現，人臉辨識系統須要面對許多不同的問題與挑戰。本研究分析目前已被提出的多種人臉辨識方法，並提出一個人臉辨識模組，稱為Enhanced-2DPCA(簡稱E-2DPCA)。將E-2DPCA方法與分析中最高辨識率的兩個方法(分別利用2DPCA與DCT coefficient)比較，結果顯示E-2DPCA的平均辨識率雖然比另外兩個方法的平均辨識率還高，三種方法對不同測試影像各有不同優劣表現，因而我們結合三種方法，利用權重式投票法取得最後辨識結果。最後結果顯示，結合的方法可更進一步改進平均辨識率。

關鍵字

人臉辨識；主成份分析；離散餘弦轉換；權重投票法；空間域；頻率域；基因演算法

並列摘要

Face Recognition is an important topic in the field of pattern recognition. This technology has a variety of applications including entrance guard control, personal service system, criminal verification, and security verification of finance. Our research focuses on the development of a human face recognition system. To correctly identify a human in an image is a challenge due to various possible factors, including different light conditions, change of haircut, variation of face expression, and different aspects of the face. We analyzed several existing face recognition techniques and found that each of them performs well over some sets of testing samples but poorly over some other sets. This motivated us to combine some different techniques to construct a better face recognition system. First, we proposed a new module, called Enhanced-2DPCA (or E-2DPCA). The accuracy of E-2DPCA is better than all the techniques we have analyzed. We chose the best two from those analyzed and compared them with our proposed E-2DPCA module, and found that although the E-2DPCA module outperforms the other two modules, each of the three modules behaves better than others over some set of samples. Thus we combine the three modules and apply weighted voting scheme to choose the recognition result from the results given by the three modules. Experimental results show that the integrated system can further improve the recognition rate.

並列關鍵字

Face Recognition ； Principle Component Analysis(PCA) ； Two Dimension Principle Component Analysis(2DPCA) ； Discrete Cosine Transformation(DCT) ； Weight-Voting ； spatial domain ； frequency domain

參考文獻

2. P. Yampri, C. Pintavirooj, S. Daochai, and S. Teartulakarn, “White Blood Cell Classification based on the Combination of Eigen Cell and Parametric Feature Detection,” Papers of Technical Meeting on Medical and Biological Engineering, Vol. 6, No. 95-115, pp. 1-4, 2006.

3. Kazuma Shigenari, Fumihiko Sakaue, and Takeshi Shakunaga, “Decomposition and Virtualization of Eigenface for Face Recognition under Various Lighting Conditions,” Systems and Computers in Japan, Vol. 36, No. 1, pp. 25-34, 2005.

5. Kwan-Ho Lin, Kin-Man Lam, and Wan-Chi Siu, “Spatially Eigen-weighted Hausdorff Distances for Human Face Recognition,” Pattern Recognition 36, No. 8, pp. 1827-1834, August 2003.

7. Xiaopeng Hong, Hongxun Yao, Yuqi Wan, and Rong Chen, “A PCA based Visiual DCT Feature Extraction Method for Lip-Reading,” in Proceeding of the 2006 International Conference on Intelligent Information Hiding and Multimedia Signal Proceeding (IIH-MSP’06), pp. 321-326, December 2006.

8. Jian Yang, David Zhang, Alejandro F. Frangi, and Jing-yu Yang, “Two-Dimensional PCA: A New Approach to Appearance-Based Face Representation and Recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 26, No. 1, pp. 131-137, January 2004.

國際替代計量

應用二維主成分分析之人臉辨識

全文下載

主題瀏覽