LSO-AdaBoost Based Face Detection for IP-CAM Video

Abstract:

The IP-CAM plays a major role in the context of digital video surveillance systems. The function of face detection can add extra value and can contribute towards an intelligent video surveillance system. The cascaded AdaBoost-based face detection system proposed by Viola can support real-time detection with a high detection rate. The performance of the Alt2 cascade (from OpenCV) in an IP-CAM video is worse than that with regard to static images because the training data set in the Alt2 cannot consider the localized characters in the special IP-CAM video. Therefore, this study presents an enhanced training method using the Adaboost algorithm which is capable of obtaining the localized sampling optimum (LSO) from a local IP-CAM video. In addition, we use an improved motion detection algorithm that cooperates with the former face detector to speed up processing time and achieve a better detection rate on video-rate processing speed. The proposed solution has been developed around the cascaded AdaBoost approach, using the open-CV library, with a LSO from a local IP-CAM video. An efficient motion detection model is adopted for practical applications. The overall system performance using 30% local samples can be improved to a 97.9% detection rate and reduce detection time by 54.5% with regard to the Alt2 cascade.

• Periodical: Applied Mechanics and Materials (Volumes 284-287)
• Pages: 3543-3548
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.284-287.3543
• Online since: January 2013
• Authors: Chuang Jan Chang, Shu Lin Hwang
• Keywords: Adaboost, Detection, IP-CAM, Video

References

[1] Chen D. S., Liu Z. K. A Survey of Skin Color Detection. Chinese Journal of Computers (2005).

[2] Huang Y. T., Chiang C. C. (2002). Real-Time Detection of Multiple Faces Using Rule-based Methods. Dong-Hwa University Taiwan (2002).

[3] Yoav Freund and Robert E. Schapire., A decision-theoretic generalization of on-line learning and an application to boosting, Computational Learning Theory: Eurocolt '95, (1995) pages 23–37.

DOI: 10.1007/3-540-59119-2_166

[4] Viola P., Jones M. J. Rapid Object Detection using a Boosted Cascade of Simple Features. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (2001).

DOI: 10.1109/cvpr.2001.990517

[5] Kinh Tieu and Paul Viola., Boosting Image Retrieval, In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2000).

DOI: 10.1109/cvpr.2000.855824

[6] Rainer Lienhart, Alexander Kuranov, and Vadim Pisarevsky, Empirical Analysis of Detection Cascades of Boosted Classifiers for Rapid Object Detection, Intel Corporation, Santa Clara, CA 95052, USA MRL Technical Report, (2002) December.

DOI: 10.1007/978-3-540-45243-0_39

[7] Freund Y., Schapire R. E. A decision-theoretic generalization of online learning and an application to Boosting. Journal of Computer and System Sciences (1997).

DOI: 10.1006/jcss.1997.1504

[8] OpenCV Reference Manual v2. 1, (2010) http: /www. opencv. org. cn/index. php.