即時監控系統之嵌入式軟體平台設計

The Design of Real-Time Surveillance with the Embedded Software Platform

高志忠

指導教授 ： 高文忠 ； 黃奇武

國立臺灣師範大學/科技學院/工業教育學系/碩士(2005年)

摘要

人臉偵測與人臉辨認在監視系統、智慧型人機介面與安全資訊擷取上面非常有用。有非常多傑出的論文提出演算法解決人臉偵測與人臉辨認的問題。然而，非常少的論文考慮關於整合曝光控制、影像擷取、人臉追蹤與有效率人臉辨認的系統並且實現於行動的數位相機系統上。
本論文提出一個完整的監視系統並且實現於一個商業的數位相機產品上，本方法一開始先針對使用膚色機率模型偵測到的人臉進行調整曝光控制來確保臉部的曝光控制正確。接下來抽取出眼睛、鼻子與嘴唇的可能位置，然後正確的人臉位置就可以藉由特徵間相對位置的模組來確認。最後，被切割出的人臉區域使用DCT係數的執行人臉辨認並且使用支持向量機來分類。實驗結果顯示提出的系統很穩定並且可以達到及時的應用。

關鍵字

人臉偵測 ； 人臉辨認 ； 支援向量機 ； 嵌入式系統 ； 數位相機

並列摘要

Face detection as well as recognition is particular useful for surveillance systems, intelligent human-computer interface and security information access. Many excellent algorithms have been proposed to solve the problems of face detection as well as recognition. However, very few of them considered the integration issues from exposure control, image capture, face tracking and efficient recognition for realizing such a system in portable camera system.

In this thesis, we propose a complete surveillance which is implemented on a commercial digital camera. The system first adaptively control the exposure based on the locations of human face which is detected by probability skin color model to ensure the right exposure of human face. The next step is to extract the possible locations of eyes, noses and lips. Then the right location of human face can be determined with relative location models. Finally the human faces are recognized based on DCT coefficients of the segmented image areas and classified with support vector machines. The experimental results show that the proposed system is stable and fast enough for real-time applications.

並列關鍵字

Face Detection ； Face Recognition ； Support Vector Machine ； Embedded System ； Digital Camera

參考文獻

[1] M. Yang, “Detecting faces in images,” IEEE Trans. Pattern Anal. Machine Intell., vol. 24, no. 1, pp. 34-58, Jan. 2002.

[2] H. A. Rowley, “Neural network-based face detection,” IEEE Trans. Pattern Anal. Machine Intell., vol. 20, no. 1, pp. 23-38, Jan. 1998.

[3] C. Garcia and G. Tziritas, “Face detection using quantized skin color regions merging and wavelet packet analysis,” IEEE Trans. Meltimedia, vol. 1, no. 3, pp. 264-277, Sep. 1999.

[4] F. Tsalakanidou, S. Malassiotis and M. G. Strintzis, “Face localization and authentication using color and depth images,” IEEE Trans. Image processing, vol. 14, no. 2, pp. 152-168, Feb. 2005.

[5] K. K. Sung and T. Poggio, “Example-based learning for view-based human face detection,” IEEE Trans. Pattern Anal. Machine Intell., vol. 20, no. 1, pp. 39-51, Jan. 1998.