一种改进的CAMShift跟踪算法及人脸检测框架

计算机工程与科学

一种改进的CAMShift跟踪算法及人脸检测框架

杨超，蔡晓东，王丽娟，朱利伟

（桂林电子科技大学信息与通信学院，广西桂林 541004）

收稿日期:2015-06-18 修回日期:2015-07-17 出版日期:2016-09-25 发布日期:2016-09-25
基金资助:
桂林电子科技大学研究生教育创新计划资助项目(YJCXS201521)

An improved CAMShift tracking algorithm and a face detection framework

YANG Chao,CAI Xiao-dong,WANG Li-juan,ZHU Li-wei

（School of Information and Communication, Guilin University of Electronic Technology, Guilin 541004, China）

Received:2015-06-18 Revised:2015-07-17 Online:2016-09-25 Published:2016-09-25

摘要/Abstract

摘要：

为充分利用人脸视频图像序列中的时空信息，获得更加准确的人脸比对图像序列，提出一种结合人脸跟踪的人脸检测框架。使用简单快速的正面人脸检测算法对人脸视频图像序列进行检测，用检测的结果对人脸跟踪算法进行初始化及校验和调整。为解决CAMShift跟踪算法容易受类肤色区域影响而导致提取到的人脸区域存在冗余信息的问题，提出一种改进的CAMShift-KLT算法。该算法利用兴趣点跟踪人脸图像的边缘，达到准确获取人脸比对图像的目的。实验结果表明，与CAMShift算法相比，CAMShift-KLT算法获取的人脸区域更精准，同时具有较小的跟踪偏移距离、较大的跟踪命中率和更高的跟踪有效性。与对比算法相比，CAMShift-KLT算法能够获得与理想的人脸区域更加一致的跟踪区域。

关键词: 人脸跟踪, 人脸检测, CAMShift, 视频, KLT

Abstract:

To make full use of human face videos in temporal and spatial domains and obtain accurate face image sequence for face alignment, we propose a face detection framework in combination with a face tracking algorithm. We use the simple and fast frontal face detection algorithm to detect face images from videos, and then employ the detection results to initialize, check and adjust the face tracking results. However, the CAMShift tracking algorithm is vulnerable to the negative impact of skin color areas, which leads to more redundant information. To solve this problem, we propose an improved CAMShift-KLT algorithm, which utilizes interest points to trace the edge of the face image and acquires accurate face alignment images. Experimental results show that compared with the CAMShift algorithm, the proposed algorithm can obtain more accurate face regions, and meanwhile, it has smaller tracking migration distance, higher hit rate of tracking and tracking effectiveness. Besides, compared with the contrast algorithm, the improved CAMShift-KLT algorithm can better track the face region.

Key words: face tracking, face detection, CAMShift, video, KLT

杨超，蔡晓东，王丽娟，朱利伟. 一种改进的CAMShift跟踪算法及人脸检测框架[J]. 计算机工程与科学.

YANG Chao,CAI Xiao-dong,WANG Li-juan,ZHU Li-wei. An improved CAMShift tracking algorithm and a face detection framework [J]. Computer Engineering & Science.

参考文献

［1］Li Yan,Wang Ling.Based on the color of skin of face detection and feature of Haar variance ［J］.Computer Engineering & Science, 2015,37(1):146-151.(in Chinese)
［2］Zhao Nan-nan.Based on AdaBoost algorithm and the skin color model profile face detection ［J］.Computer Engineering & Science,2011,33(5):85-90.(in Chinese)
［3］Huang Yu-xin,Huang Shan,Zhang Hong-bin.Based on the color of skin and organs positioning real-time face detection ［J］.Computer Engineering & Science,2014,36(5):938-940.(in Chinese)
［4］Xu Xue-mei, Huang Zheng-yu, Li Li-xian, et al. A method based on HSV color segmentation and face detection technology of fuzzy cascade classifier ［J］.Journal of Computer Applications and Software, 2014,31(4):197-200.(in Chinese)
［5］Li Lan, Zhu Lin-jie, Zhang Yun.Skin color and gradient complementary face detection research ［J］.Journal of Computer Applications and Software, 2013,30(2):280-282.(in Chinese)
［6］Kalal Z,Mikolajczyk K, Matas J. Face-TLD:Tracking-Learning-Detection applied to faces［C］∥Proc of the 17th IEEE International Conference on Image Processing(ICIP),2010:3789-3792.
［7］Wu X,Li L,Lai J,et al.A framework of face tracking with classification using CAMShift-C and LBP［C］∥Proc of the 5th IEEE International Conference on Image and Graphics,2009:217-222.
［8］Zahir N B,Samad R,Mustafa M.Initial experimental results of real-time variant pose face detection and tracking system［C］∥Proc of IEEE International Conference on Signal and Image Processing Applications (ICSIPA),2013:264-268.
［9］Li X,Zhou X. Automatic real-time face detection and tracking based on space-temporal mutual feedback for video sequence［C］∥Proc of IEEE International Conference on Audio,Language and Image Processing,2008:1650-1654.
［10］Ishil I,Ichida H,Takaki T.GPU-based face tracking at 500 fps［C］∥Proc of the 18th IEEE International Conference on Image Processing (ICIP),2011:557-560.
［11］Shi J,Tomasi C.Good features to track［C］∥Proc of IEEE Computer Society Conference on Computer Vision and Pattern Recognition,1994:596-600.
［12］Li J,Zhang Y. Learning SURF cascade for fast and accurate object detection［C］∥Proc of IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2013:3468-3475.
［13］Sanderson C, Lovell B C. Multi-region probabilistic histograms for robust and scalable identity inference［C］∥Proc of ICB09,2009:199-208.
［14］Gorodnichy D.Video-based framework for face recognition in video［C］∥ Proc of the 2nd Workshop on Face Processing in Video (FPiV'05) in Proceedings of 2nd Canadian Conference on Computer and Robot Vision (CRV’05),2005:330-338.
［15］Zhang Kai-hua,Zhang Lei,Yang Ming-hsuan.Real-time compressive tracking［C］∥Proc of European Conference on Computer Vision (ECCV),2012:864-877.
附中文参考文献:
［1］李燕,王玲.基于肤色和Haar方差特征的人脸检测［J］.计算机工程与科学,2015,37(1):146-151.
［2］赵男男.基于AdaBoost算法与肤色模型的多姿态人脸检测［J］.计算机工程与科学,2011,33(5):85-90.
［3］黄禹馨,黄山,张洪斌.基于肤色和器官定位的实时人脸检测［J］.计算机工程与科学,2014,36(5):938-940.
［4］许雪梅,黄征宇,李丽娴,等.一种基于HSV颜色分割和模糊级联分类器的人脸检测技术［J］.计算机应用与软件,2014,31(4):197-200.
［5］李岚,朱林杰,张云.肤色与梯度互补的人脸检测研究［J］.计算机应用与软件,2013,30(2):280-282.

[1]	崔馨月, 蒋林, 杨坤, 惠超, 胡传瞻, 赵静. 一种HEVC帧内预测算法的动态自重构实现方法[J]. 计算机工程与科学, 2022, 44(12): 2120-2127.
[2]	何平, 李刚, 李慧斌, . 基于深度学习的视频异常检测方法综述[J]. 计算机工程与科学, 2022, 44(09): 1620-1629.
[3]	刘紫燕, 袁磊, 朱明成, 马珊珊. 融合改进通道和层剪枝的口罩人脸检测[J]. 计算机工程与科学, 2022, 44(03): 463-470.
[4]	张元鸣, 虞家睿, 陆佳炜, 高飞, 肖刚. 基于Spark Streaming的视频大数据并行处理方法[J]. 计算机工程与科学, 2021, 43(10): 1736-1743.
[5]	黄展原, 李兵, 李庚浩, . 基于视频和人体姿态估计的老年人摔倒监测研究[J]. 计算机工程与科学, 2021, 43(05): 883-890.
[6]	程小兰, 蒋从锋, 欧东阳, 任永坚, 张纪林, 万健, . 基于帧缓冲队列的边缘视频处理加速方法[J]. 计算机工程与科学, 2021, 43(04): 603-613.
[7]	曾上游, 贾小硕, 李文惠. 基于双线程LSTM在线更新的视频追踪算法[J]. 计算机工程与科学, 2021, 43(03): 480-485.
[8]	杨坤, 蒋林, 谢晓燕, 邓军勇, 刘新闯, 胡传瞻. HEVC中率失真优化算法的动态可重构实现[J]. 计算机工程与科学, 2021, 43(02): 354-361.
[9]	王思齐, 胡婧韬, 余广, 祝恩, 蔡志平. 智能视频异常事件检测方法综述[J]. 计算机工程与科学, 2020, 42(08): 1393-1405.
[10]	熊炜1,2，王传胜1，管来福1，童磊1，刘敏1，曾春艳1. 基于特征跟踪和网格路径运动的视频稳像算法[J]. 计算机工程与科学, 2020, 42(05): 843-850.
[11]	孙明华，杨媛，李渊博. 基于深度可分离卷积的地铁隧道巡检视频分析[J]. 计算机工程与科学, 2020, 42(04): 691-698.
[12]	熊炜1,2，王传胜1，李利荣1，刘敏1，曾春艳1. 结合光流法和卡尔曼滤波的视频稳像算法[J]. 计算机工程与科学, 2020, 42(03): 493-499.
[13]	夏珺1，钱磊2，严伟3，柴志雷1. 基于FPGA的HEVC后处理CNN硬件加速器研究[J]. 计算机工程与科学, 2018, 40(12): 2126-2132.
[14]	陶祝,刘正熙,熊运余,李征. 基于深度神经网络的行人头部检测[J]. 计算机工程与科学, 2018, 40(08): 1475-1481.
[15]	邱亮1，夏慧明1，储久良2. 基于时域梯度相似度的视频质量评价模型[J]. 计算机工程与科学, 2018, 40(04): 707-711.