تشخیص زاویه قرارگیری شخص در تصویر با استفاده از اطلاعات کانتوری
محورهای موضوعی : مهندسی برق و کامپیوتر
1 - دانشگاه صنعتی شاهرود
2 - دانشگاه صنعتی شاهرود
کلید واژه: اطلاعات کانتوری شخص خوشهبندی سلسلهمراتبی سیستمهای نظارتی,
چکیده مقاله :
در بسیاری از کاربردها از جمله کاربردهای نظارتی، حالت و زاویه قرارگیری شخص نسبت به دوربین از اطلاعات مهم میباشد. این اطلاعات میتواند در ارزیابی رفتاری شخص مورد استفاده قرار گیرد. از دشواریهای این تشخیص میتوان به کیفیت پایین دوربینهای نظارتی، نویز و پسزمینههای پیچیده در تصویر اشاره نمود. در روشهای موجود برای تشخیص زاویه از ویژگیهایی نظیر هیستوگرام گرادیانهای جهتی استفاده میشود. در این توصیفگر محاسبه هیستوگرامها بر اساس نواحی محلی انجام میگیرد که دارای نقاط ضعفی در تشخیص زاویه میباشد. یکی از اطلاعات مفید که میتواند در تشخیص زاویه مورد استفاده قرار گیرد لبههای احاطهکننده یک شیء در تصویر است که بدان کانتور شیء اطلاق میگردد. در این مقاله تعمیمی از کانتور ارائه میشود که به کمک آن میتوان به صورت سلسلهمراتبی تخمینی از کانتور محاطشده به تصویر شخص را ارائه داد. این کانتورها از روی یک مدل سهبعدی انسان تولید میشوند. کانتور تخمین زده شده همانند یک ویژگی سطح بالا در کنار ویژگیهای سطح پایین نظیر هیستوگرام گرادیانهای جهتی به عنوان ویژگی نهایی در نظر گرفته میشود. در تولید این ویژگی از ترکیب خطی چندین نوع کانتور مربوط به بخشهای مختلف بدن استفاده شده است. به منظور نشاندادن تأثیر ویژگی جدید در تشخیص زاویه، کلاسبند ماشین بردار پشتیبان با استفاده از ترکیب دو ویژگی بالا آموزش داده شده و سپس بر روی مجموعه دادههای VIPeR مورد ارزیابی قرار گرفته است. نتایج آزمایشات انجامشده نشان میدهد که استفاده از ویژگی ارائهشده، دقت تشخیص زاویه را حدود 4% بهبود میبخشد.
Pose and orientation of a person relative to the camera are the important and useful information in many applications, including surveillance systems. This information can be used in the behavior analysis of the person. Low quality of the recorded surveillance images, noisy data and cluttered backgrounds are some of the difficulties in this task. In the existing methods, histogram of orientation gradient (HOG) is used to estimate the orientation. The local properties of HOG is a weakness for orientation estimation. The edge surrounding the object, namely contour, is a useful information for orientation estimation. In this paper we present a general form of a contour. This hyper contour helps us to find the best contour which is matched to image of the person in a hierarchical fashion. These contours generated from a human 3D model. The matched contour as a high-level feature is combined with the low-level feature such as HOG, and considered as the final feature. The proposed feature is a linear combination of several types of contours with respect to different regions of the body. To show the impact of the proposed feature on orientation estimation, a support vector machine is trained on a hybrid feature space and then is evaluated on VIPeR dataset. The experimental results show that the accuracy of the orientation estimation is improved about 4% by using the extended feature.
[1] K. Smith, S. O. Ba, J. M. Odobez, and D. Gatica-Perez, "Tracking the visual focus of attention for a varying number of wandering people," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 30, no. 7, pp. 1212-1229, Jul. 2008.
[2] ع. سبطی و ح. حسنپور، "بهبود الگوریتم SDALF در بازشناسی انسان با بهرهگیری از اطلاعات زاویه شخص،" بیست و سومین کنفرانس مهندسی برق ایران، صص. 894-890، اردیبهشت 1394.
[3] G. Fanelli, J. Gall, and L. Van Gool, "Real time head pose estimation with random regression forests," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'11, pp. 617-624, Jun. 2011.
[4] D. Baltieri, R. Vezzani, and R. Cucchiara, "Sarc3d: a new 3d body model for people tracking and re-identification," in Proc. Int.Conf. onImage Analysisand Processing ICIAP'11, pp. 197-206, Sept. 2011.
[5] L. Bourdev and J. Malik, "Poselets: body part detectors trained using 3d human pose annotations," in Proc. IEEE 12th Int. Conf. on Computer Vision, pp. 1365-1372, Sep. 2009.
[6] D. Tosato, M. Spera, M. Cristani, and V. Murino, "Characterizing humans on riemannian manifolds," , IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 35, no. 8, pp. 1972-1984, Aug. 2013.
[7] J. Tao and R. Klette, "Integrated pedestrian and direction classification using a random decision forest," in Proc. IEEE Int. Conf. on Computer Vision Workshops, ICCVW'13, , pp. 230-237, Dec. 2013.
[8] M. Andriluka, S. Roth, and B. Schiele, "Pictorial structures revisited: people detection and articulated pose estimation," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'09, pp. 1014-1021, Jun. 2009.
[9] M. Andriluka, S. Roth, and B. Schiele, "Monocular 3d pose estimation and tracking by detection," in Proc., IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'10, pp. 623-630, Jun. 2010.
[10] D. M. Gavrila and S. Munder, "Multi-cue pedestrian detection and tracking from a moving vehicle," International J. of Computer Vision, vol. 73, no. 1, pp. 41-59, Jun. 2007.
[11] Y. Mu, S. Yan, Y. Liu, T. Huang, and B. Zhou, "Discriminative local binary patterns for human detection in personal album," in Proc., IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'08, 8 pp, 23-28 Jun. 2008.
[12] O. Tuzel, F. Porikli, and P. Meer, "Pedestrian detection via classification on riemannian manifolds," IEEE Trans. on Pattern Analysis and Machine Intelligencevol. 30, no. 10, pp. 1713-1727, Oct. 2008.
[13] P. F. Felzenszwalb, R. B. Girshick, D. McAllester, and D. Ramanan, "Object detection with discriminatively trained part-based models," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 32, no. 9, pp. 1627-1645, Sept. 2010.
[14] Q. Zhu, M. C. Yeh, K. T. Cheng, and S. Avidan, "Fast human detection using a cascade of histograms of oriented gradients," in Proc. IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, CVPR'06, vol. 2, pp. 1491-1498, Jun. 2006.
[15] N. Dalal and B. Triggs, "Histograms of oriented gradients for human detection," in Proc. IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, CVPR'05, vol. 1, pp. 886-893, Jun. 2005.
[16] ع. سبطی و ح. حسنپور، "بازشناسی انسان در سيستمهای نظارت ويدئويی،" مجله محاسبات نرم دانشگاه کاشان، جلد سوم، شماره اول، صص. 81-62، تابستان 1393.
[17] D. Huang, M. Storer, F. De la Torre, and H. Bischof, "Supervised local subspace learning for continuous head pose estimation," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'11, pp. 2921-2928, Jun. 2011.
[18] E. Murphy-Chutorian and M. M. Trivedi, "Head pose estimation in computer vision: a survey," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 31, no. 4, pp. 607-626, Apr. 2009.
[19] D. Tran and D. Forsyth, "Improved human parsing with a full relational model," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'10, pt. 4, , pp. 227-240, Sep.2010.
[20] M. Enzweiler and D. M. Gavrila, "Integrated pedestrian classification and orientation estimation," in Proc. IEEE Conf. Computer Vision and Pattern Recognition, CVPR'10, pp. 982-989, Jun. 2010.
[21] I. Chavel, Riemannian Geometry: A Modern Introduction, Cambridge University Press, vol. 98, Apr. 2006.
[22] N. F. Troje, "Decomposing biological motion: A framework for analysis and synthesis of human gait patterns," Journal of Vision, vol. 2, no. 5, pp. 371-387, Sep. 2002.
[23] R. Kimmel, N. Kiryati, and A. M. Bruckstein, "Sub-pixel distance maps and weighted distance transforms," J. of Mathematical Imaging and Vision, vol. 6, no. 2-3, pp. 223-233, Jun. 1996.
[24] M. P. Dubuisson and A. K. Jain, "A modified Hausdorff distance for object matching," in Proc. of the 12th IAPR Int. Conf. IEEE Pattern Recognition, Computer Vision & Image Processing, vol. 1, pp. 566-568, Oct. 1994.
[25] L. Rokach and O. Maimon, "Clustering methods," in Data Mining and Knowledge Discovery Handbook, Springer, pp. 321-352, 2005.
[26] D. Gray, S. Brennan, and H. Tao, "Evaluating appearance models for recognition, reacquisition, and tracking," in Proc. IEEE International Workshop on Performance Evaluation of Tracking and Surveillance, vol. 3, pp. 41-47, Oct. 2007.
[27] S. H. Cha, "Comprehensive Survey on Distance/Similarity Measures between Probability Density Functions, " Int. J. of Mathematical Models and Methods in Applied Sciences, vol. 1, no. 4, pp. 300-307, 2007.
[28] C. C. Chang, and C. J. Lin, "LIBSVM: a library for support vector machines," ACM Trans. on Intelligent Systems and Technology, vol. 2, no. 3, Article no. 27, Apr. 2011.
[29] P. Felzenszwalb and D. Huttenlocher, Distance Transforms of Sampled Functions, Cornell University, 2004.