¹Beijing Key Laboratory of Information Service Engineering, Beijing Union University, Beijing 100101, China.
²Institute of Computer Technology, Beijing Union University, Beijing 100101, China.
³Beijing Advanced Innovation Center for Imaging Technology, Capital Normal University, Beijing 100101, China.

---

Abstract—Face detection is the basis for research topics such as face recognition, facial expression recognition and face attribute analysis and it plays a crucial role in the field of computer vision. Traditional methods are difficult to meet the needs of robust real-time multi face detection because of some influencing factors such as head pose, image scene, illumination condition and so on. In this paper, we introduce an intelligent vision algorithm that is able to detect human face from complex scene and filter out all the non-face but face-like images. The human face is detected in real-time environment using the approach called Adaboost-based Haar-Cascade Classifier, and the real human face detection is improved from single-face detection to multi-face detection. In addition, variable head poses are taken into account, such as pitch, roll, yaw, etc. Furthermore, the real-time experiments proved the effectiveness and robustness of the algorithm for human detection we have proposed.

Index Terms—Face detection, adaboost algorithm, strong classifier, real-time, multi-pose.

[PDF]

Cite: Xin-Chao Zhao, Jia-Zheng Yuan, Hong-Zhe Liu, Jian-She Zhou, "Improved AdaBoost Algorithm for Robust Real-Time Multi-face Detection," Journal of Software vol. 12, no. 1, pp. 53-61, 2017.