基于多目标跟踪的非头肩人脸跟踪研究
|
摘要
人脸检测与跟踪是计算机视觉方面的一个重要和前沿的研究课题。本文针对复杂背景条件下的非头肩人脸视频序列的检测和跟踪技术进行了深入的研究。根据非头肩人脸视频序列的特点，本文采用在目标跟踪的基础上进行人脸检测，从而实现人脸跟踪的解决方法，涉及到目标检测、人脸检测、目标跟踪等相关课题，经实验证明适用于非头肩人脸视频序列。具体内容为：采用背景提取和帧间差分相结合的算法检测出运动目标，然后使用Kalman预测和SSDA目标匹配算法在视频序列中完成实时目标跟踪。人脸检测在目标跟踪的基础上完成，在跟踪到的人体图像上首先采用人脸模板相似度匹配进行粗筛选，经过粗筛选的疑似人脸图像再经一阶Harr小波变换提取特征，最后通过支持向量机进行分类。本文实验程序采用Visual C++实现。
Automatic face detection and tracking is an important and front-line subject in computer vision domain. In this paper, we have made a deep study on the face detection and tracking in the non-head and shoulder image sequence. According to the characteristic of non-head and shoulder image sequence, we present an approach that the face detection is based on the object detection and tracking. This approach relates to the fields of object detection object tracking and face detection, which is proved effective. Though using the method of both multi-frame difference and backgroud estimation to sequence images, the detection result is acquired. Then the Kalman filter is used to predict the moving object. Combined with the improved SSDA match method, we can track the moving object exactly on real time. We detect face in the part of the tracked body image. The face-template match is used to get the probably face area. Then after doing harr-wavelet transformation to the probably face area, we use support vector machine to distinguish face accurately. All the programs for testing are written in Visual C++.
Automatic face detection and tracking is an important and front-line subject in computer vision domain. In this paper, we have made a deep study on the face detection and tracking in the non-head and shoulder image sequence. According to the characteristic of non-head and shoulder image sequence, we present an approach that the face detection is based on the object detection and tracking. This approach relates to the fields of object detection object tracking and face detection, which is proved effective. Though using the method of both multi-frame difference and backgroud estimation to sequence images, the detection result is acquired. Then the Kalman filter is used to predict the moving object. Combined with the improved SSDA match method, we can track the moving object exactly on real time. We detect face in the part of the tracked body image. The face-template match is used to get the probably face area. Then after doing harr-wavelet transformation to the probably face area, we use support vector machine to distinguish face accurately. All the programs for testing are written in Visual C++.
引文
1. J.Barron, D.Fleet, and S.Beauchemin, Performance of Optical Flow Techniques. International Journal of Computer Vision, 1994.12(1): p. 42-77.
2. N.Ohta and K.Kanatani, Moving object detection from optical flow without empirical thresholds. IEICE Transaction on Information and System, 1998. E81-D(2): p. 243-245.
3. S.Yamamoto, et al., Realtime multiple object tracking based on optical flows. IEEE International Conference on Computer Vision, 1995.3: p. 2328-2333.
4. C.Wren, et al., Pfinder:Real-time Tracking of the Human Body. IEEE Trans.PAMI, 1997.19(7): p. 780-785.
5. I.Haritaoglu, D.Harwood, and L.S.David, Real-Time Surveillance of People and Their Activities. IEEE Trans.PAMI, 2000.22(8): p. 809-830.
6. K.Toyma, et al., Wallflower:Principle and practice of backgroud maintenace. International Conference of Computer Vision, 1999: p. 255-261.
7. Y.Z.Hsu, H.H.Nagel, and G.Rebers, New Likehood Test Methods for Change Detection in Image Sequence. Computer Vision,Graphics and Image Processing, 1984.26: p. 73-106.
8. P.L.Rosin and T.Ellis, Image Difference Threshold Strategies and Shadow Detection. In Processing of the British Machine Vision Conference, 1995: p. 347-356.
9. F.G.Meyer and P.Bouthemy, Region Based Tracking using Affine Motion Models in Long Image Sequence. Computer Vision,Graphics and Image Processing:Image Unstanding, 1994.60(2): p. 119-140.
10. Badenas, J., J.M. Sanchiz, and F. pla, Motion_based segmentation and region tracking in image sequence. Pattern Recognition, 2001.34(3): p. 661-670.
11. K-K, S. and P. T, Example-based learning for view-based human face detection. IEEE Transactions on Patten Analysis and Machine Intelligence, 1998.20(1): p. 39-51.
12. Turk, M. and A. Pentland, Egienfaces for recoginition. Journal of cognitive neuroscience, 1991.3(1): p. 71-86.
13. M, T. and P. A, Eigenfacesfor recognation. Journal of cognitive neuroscience, 1991.3(1): p. 71-86.
14. Moghaddam, B. and A. pentland, Face recognition using view-based and modular eigenspaces. Proceeding of SPIE 2257, 1994: p. 12-21.
15. Yang, G.Z., Pattern Recognition, 1994.27: p. 53-63.
16. R.Chellappa, Proceeding of IEEE, 1995.83: p. 705-740.
17. V.Covindaraju, Computer Vision, 1996.19: p. 129-146.
18. A.Tankus, PR Letters, 1997.18: p. 913-922.
19. R.Brunel and T. Poggio, Face recognition:Features versus templates. IEEE Transactions on Patten Analysis and Machine Intelligence, 1993. 15(10): p. 1042-1052.
20. A.Yullie, P.Hallinan, and D.Cohen, Feture Exaction from faces using deformable templates. International Journal of Computer Vision, 1992.8(2): p. 99-111.
21. A-k.Jain, Y. Zhong, and M-P.Dubuission-Jolly, Deformable templates models:A review. Signal Processing, 1998.71(2): p. 109-129.
22. J.Cai and A.Goshtasby, Detecting human faces in color images. Image and Vision Computing, 1999.18(1): p. 63-75.
23. R.L.Hsu, Face detection in color images. IEEE Trans.on Patten Analysis and Machine Intelligence, 2002.24(5): p. 696-706.
24. C.Papageorgiou, T.Poggio, and O.renM, A General Framework for Object Detection. Proc.Sixth IEEE Int'l Conf.Computer Vision, 1998.
25. G.Amara, 2004. An Introduction to Wavelets.
26. Vapnik and V.N, The nature of statistical learning theory. Springer-Verlag, 1995.
27．边肇祺，模式识别．2000：清华大学出版社．
28．张学工，关于统计学习理论与支持向量机．自动化学报，2000．26(1)．
29．陈朝阳 and 张桂林，基于图像对称差分运算的运动小目标检测方法．华中理工大学学报，1998．26(9)．
30．李凌娟 and 贾振堂，一种鲁棒的视频分割算法．中国图像图像学报，2002．7(11)．
31. Jung, Y.-K. and Yo-Sung, Traffic parameter extraction using video-based vehicle tracking. IEEE International Conference on.5-8, 1999: p. 764-769.
32. Peihong, L., D. liya, and L. jilin, A video-based traffic information extraction system. Intelligent Vehicles Symposium, proceedings. IEEE, June 9-11, 2003: p. 528-532.
33．郑江滨，智能视频监视研究．西北工业大学博士学位论文，2002．
34．肖敬若，复杂背景下的实时目标跟踪技术研究．西北工业大学硕士学位论文，2006。
35. Otsu, A threshold selection method from gray-lever histogram. IEEE Trans. on SMC, 1979. 9(1): p. 62-66.
36．王伟 and 孙敏，一种交通场景下运动目标检测的新方法．计算机应用与软件，2005．22(4)：p．90-92．
37．张贤达，现代信号处理．2002：清华大学出版社．
38. I.Haritaoglu and D.Harwood, Active outdoor survellance. Proceedings International Confenrence on Image Analysis and processing, 1999: p. 1096-1099.
39. Tekalp, M., Digital Video Processing. Prentice Hall,inc 19.
40. M.Isard and A.Blake, Condensation-conditional density propagation for visual tracking. International Journal of Computer Vision, 1998.29(1): p. 5-28.
41. T.J.chain, Multiple Hypothesis Approach to Figure Tracking. In Conference on Computer Vision and Pattern Recognition,Fort Collins,Colarado, J.M.Rehg: p. 23-25.
42．姜锦锋，红外图像的目标检测、识别与跟踪技术．西北工业大学硕士学位论文，2004：p．65—68．
43. D.I.Bamea, A class of algorithm for digital image registration. IEEE Trans Computer 1972. C-21: p. 176-186.
44．沈庭芝 and 方子文，数字图像处理及模式识别．1998：北京理工大学出版社．
45. zhenhua, H. and Z. peiying, An improved SSDA applied in Target tracking. Pattern Recognition,IEEE International Conference on, 1998. 2: p. 767-769.
46. Hatabu, A., T. Miyazaki, and I. kuroda, Optimization of decision_timing for early termination of SSDA-based block matching, Acoustics, Speech and Signal Processing. IEEE International Conference on, 2003.2.
47．何斌，Visual C++数字图像处理(第二版)．2002．
48. W.Sweldens, The lifting Scheme:A New philosophy in biorthogonal wavelet constructions[A], proceeding of SPIE on Wavelet Application in Signal and Image ProcessingⅢ[C], 1995: p. 68-79.
49. Sung, K.K. and T.Poggio, Example-based Learing for View-Based Human Face Detection. Technical Report AI Memo1521.Massachusetts Inst.of Technology AI Lab, 1994.
50．陈杰，陈熙霖，and 高文，基于遗传算法重采样的人脸样本扩张．Journal of Software，2005．16(11)：p．1894-1901．
51. J.C.Platt, Fast training of support vector machines using sequential minimal optimization[A],Advances in Kernal Methods:Support Vector Learing. Cambridge.MA:MIT Press, 1999: p. 185-208.
52. S.S.Keerrthi and S.K.Shevade, Improvements to Platt's SMO algorithm for SVM classifier design[J]. Neual Computation, 2001.13(3): p. 637-649.
53．李鸿一，基于支持向量机的人脸检测。大连理工大学硕士学位论文，2005．
54．侯格贤，目标跟踪系统中相关技术的研究(博士后学位论文)。中国科学院自动化研究所博士后学位论文，2003．
55．吴培景，基于SSDA算法改进的运动目标跟踪。复旦大学硕士学位论文，2005．
56．朱麾，基于支持向量机的人脸检测的研究，中国科学技术大学硕士学位论文，2004．
2. N.Ohta and K.Kanatani, Moving object detection from optical flow without empirical thresholds. IEICE Transaction on Information and System, 1998. E81-D(2): p. 243-245.
3. S.Yamamoto, et al., Realtime multiple object tracking based on optical flows. IEEE International Conference on Computer Vision, 1995.3: p. 2328-2333.
4. C.Wren, et al., Pfinder:Real-time Tracking of the Human Body. IEEE Trans.PAMI, 1997.19(7): p. 780-785.
5. I.Haritaoglu, D.Harwood, and L.S.David, Real-Time Surveillance of People and Their Activities. IEEE Trans.PAMI, 2000.22(8): p. 809-830.
6. K.Toyma, et al., Wallflower:Principle and practice of backgroud maintenace. International Conference of Computer Vision, 1999: p. 255-261.
7. Y.Z.Hsu, H.H.Nagel, and G.Rebers, New Likehood Test Methods for Change Detection in Image Sequence. Computer Vision,Graphics and Image Processing, 1984.26: p. 73-106.
8. P.L.Rosin and T.Ellis, Image Difference Threshold Strategies and Shadow Detection. In Processing of the British Machine Vision Conference, 1995: p. 347-356.
9. F.G.Meyer and P.Bouthemy, Region Based Tracking using Affine Motion Models in Long Image Sequence. Computer Vision,Graphics and Image Processing:Image Unstanding, 1994.60(2): p. 119-140.
10. Badenas, J., J.M. Sanchiz, and F. pla, Motion_based segmentation and region tracking in image sequence. Pattern Recognition, 2001.34(3): p. 661-670.
11. K-K, S. and P. T, Example-based learning for view-based human face detection. IEEE Transactions on Patten Analysis and Machine Intelligence, 1998.20(1): p. 39-51.
12. Turk, M. and A. Pentland, Egienfaces for recoginition. Journal of cognitive neuroscience, 1991.3(1): p. 71-86.
13. M, T. and P. A, Eigenfacesfor recognation. Journal of cognitive neuroscience, 1991.3(1): p. 71-86.
14. Moghaddam, B. and A. pentland, Face recognition using view-based and modular eigenspaces. Proceeding of SPIE 2257, 1994: p. 12-21.
15. Yang, G.Z., Pattern Recognition, 1994.27: p. 53-63.
16. R.Chellappa, Proceeding of IEEE, 1995.83: p. 705-740.
17. V.Covindaraju, Computer Vision, 1996.19: p. 129-146.
18. A.Tankus, PR Letters, 1997.18: p. 913-922.
19. R.Brunel and T. Poggio, Face recognition:Features versus templates. IEEE Transactions on Patten Analysis and Machine Intelligence, 1993. 15(10): p. 1042-1052.
20. A.Yullie, P.Hallinan, and D.Cohen, Feture Exaction from faces using deformable templates. International Journal of Computer Vision, 1992.8(2): p. 99-111.
21. A-k.Jain, Y. Zhong, and M-P.Dubuission-Jolly, Deformable templates models:A review. Signal Processing, 1998.71(2): p. 109-129.
22. J.Cai and A.Goshtasby, Detecting human faces in color images. Image and Vision Computing, 1999.18(1): p. 63-75.
23. R.L.Hsu, Face detection in color images. IEEE Trans.on Patten Analysis and Machine Intelligence, 2002.24(5): p. 696-706.
24. C.Papageorgiou, T.Poggio, and O.renM, A General Framework for Object Detection. Proc.Sixth IEEE Int'l Conf.Computer Vision, 1998.
25. G.Amara, 2004. An Introduction to Wavelets.
26. Vapnik and V.N, The nature of statistical learning theory. Springer-Verlag, 1995.
27．边肇祺，模式识别．2000：清华大学出版社．
28．张学工，关于统计学习理论与支持向量机．自动化学报，2000．26(1)．
29．陈朝阳 and 张桂林，基于图像对称差分运算的运动小目标检测方法．华中理工大学学报，1998．26(9)．
30．李凌娟 and 贾振堂，一种鲁棒的视频分割算法．中国图像图像学报，2002．7(11)．
31. Jung, Y.-K. and Yo-Sung, Traffic parameter extraction using video-based vehicle tracking. IEEE International Conference on.5-8, 1999: p. 764-769.
32. Peihong, L., D. liya, and L. jilin, A video-based traffic information extraction system. Intelligent Vehicles Symposium, proceedings. IEEE, June 9-11, 2003: p. 528-532.
33．郑江滨，智能视频监视研究．西北工业大学博士学位论文，2002．
34．肖敬若，复杂背景下的实时目标跟踪技术研究．西北工业大学硕士学位论文，2006。
35. Otsu, A threshold selection method from gray-lever histogram. IEEE Trans. on SMC, 1979. 9(1): p. 62-66.
36．王伟 and 孙敏，一种交通场景下运动目标检测的新方法．计算机应用与软件，2005．22(4)：p．90-92．
37．张贤达，现代信号处理．2002：清华大学出版社．
38. I.Haritaoglu and D.Harwood, Active outdoor survellance. Proceedings International Confenrence on Image Analysis and processing, 1999: p. 1096-1099.
39. Tekalp, M., Digital Video Processing. Prentice Hall,inc 19.
40. M.Isard and A.Blake, Condensation-conditional density propagation for visual tracking. International Journal of Computer Vision, 1998.29(1): p. 5-28.
41. T.J.chain, Multiple Hypothesis Approach to Figure Tracking. In Conference on Computer Vision and Pattern Recognition,Fort Collins,Colarado, J.M.Rehg: p. 23-25.
42．姜锦锋，红外图像的目标检测、识别与跟踪技术．西北工业大学硕士学位论文，2004：p．65—68．
43. D.I.Bamea, A class of algorithm for digital image registration. IEEE Trans Computer 1972. C-21: p. 176-186.
44．沈庭芝 and 方子文，数字图像处理及模式识别．1998：北京理工大学出版社．
45. zhenhua, H. and Z. peiying, An improved SSDA applied in Target tracking. Pattern Recognition,IEEE International Conference on, 1998. 2: p. 767-769.
46. Hatabu, A., T. Miyazaki, and I. kuroda, Optimization of decision_timing for early termination of SSDA-based block matching, Acoustics, Speech and Signal Processing. IEEE International Conference on, 2003.2.
47．何斌，Visual C++数字图像处理(第二版)．2002．
48. W.Sweldens, The lifting Scheme:A New philosophy in biorthogonal wavelet constructions[A], proceeding of SPIE on Wavelet Application in Signal and Image ProcessingⅢ[C], 1995: p. 68-79.
49. Sung, K.K. and T.Poggio, Example-based Learing for View-Based Human Face Detection. Technical Report AI Memo1521.Massachusetts Inst.of Technology AI Lab, 1994.
50．陈杰，陈熙霖，and 高文，基于遗传算法重采样的人脸样本扩张．Journal of Software，2005．16(11)：p．1894-1901．
51. J.C.Platt, Fast training of support vector machines using sequential minimal optimization[A],Advances in Kernal Methods:Support Vector Learing. Cambridge.MA:MIT Press, 1999: p. 185-208.
52. S.S.Keerrthi and S.K.Shevade, Improvements to Platt's SMO algorithm for SVM classifier design[J]. Neual Computation, 2001.13(3): p. 637-649.
53．李鸿一，基于支持向量机的人脸检测。大连理工大学硕士学位论文，2005．
54．侯格贤，目标跟踪系统中相关技术的研究(博士后学位论文)。中国科学院自动化研究所博士后学位论文，2003．
55．吴培景，基于SSDA算法改进的运动目标跟踪。复旦大学硕士学位论文，2005．
56．朱麾，基于支持向量机的人脸检测的研究，中国科学技术大学硕士学位论文，2004．