基於分群法設計人臉辨識應用於學生出席自動記錄系統

Design of Face Recognition Based on Clustering for Student Present Recording System

簡宏任(Hung-Ren Chien)

指導教授 ： 蔡鴻旭

國立虎尾科技大學/管理學院/資訊管理研究所/碩士(2011年)

摘要

在傳統教室上課，當教師使用人工式點名，造成教師授課時間減短；但應用人臉影像辨識於課堂點名系統時，系統須要先收集每個學生人臉資訊，這可能導致耗費過多時於建置系統，因此本論文提出一個基於分群法技術應用人臉影像辨識於學生出席自動記錄系統，首先利用直方圖等化技術解決明亮度太暗或太亮問題，接著利用局部二值化樣本方法抽取邊緣特徵，在應用K個親和傳遞分群法於邊緣特徵建立K個親和傳遞分群模式，經本研究證實K個親和傳遞分群法優於其他本研究所探討之分群法，並可應用於學習者人臉分群。

關鍵字

人臉辨識 ； 直方圖等化 ； 局部二值化樣本 ； K個親和傳遞演算法 ； 分群法

並列摘要

The thesis proposes a face recognition technique based on clustering technology. It first adopts the histogram-equalization scheme to solve the high brightness problem. That is, a face image may be highly dark or bright. Subsequently, the proposed technique uses the local-binary-pattern (LBP) method to extract edge features for a face image. These edge features of an image can be formed as a feature vector for the image. It then utilizes the K-affinity propagation (K-AP) scheme to generate a K-AP clustering model for a set of feature vectors of images. Experimental results show that the performance of the proposed technique is better than that of other clustering methods under consideration here. Consequently, the technique can be applied to face-recognition applications.

並列關鍵字

Face recognition ； Local binary pattern ； Affinity Propagation ； Clustering, Retinex ； Histogram distritution

參考文獻

[1] D. Lortie, "Chicago, Ill, Schoolteacher (2nd ed)," University of Chicago Press, 2002.

[2] J. Calderhead and M. Robson, "Images of teaching: student teachers’ early conceptions of classroom practice," Teaching & Teacher Education, vol. 7, 1991.

[3] J. B. Sandra, "Pre-service teachers’ episodic memories of classroom management," Teaching and Teacher Education, 2010.

[4] J. E. Ormrod," Educational psychology: Developing learners(4th ed)," Upper Saddle River, NJ: Merrill/Prentice-Hall, 2003.

[5] M. L. Manning and K. T. Bucher , "Classroom management: Models, applications, and cases," Upper Saddle River, NJ: Merrill/Prentice Hall, 2003.