Human detection is important in many applications and has attracted significant attention over the last decade. The Histograms of Oriented Gradients (HOG) as effective local descriptors are used with binary sliding window mechanism to achieve good detection performance. However, the computation of HOG under such framework is about billion times and the pure software implementation for HOG computation is hard to meet the real-time requirement. This study proposes a hardware architecture called One-HOG accelerator operated on FPGA of Xilinx Spartan-6 LX-150T that provides an efficient way to compute HOG such that an embedded real-time platform of HW/SW co-design for application to crowd estimation and analysis is achieved. The One-HOG accelerator mainly consists of gradient module and histogram module. The gradient module is for computing gradient magnitude and orientation; histogram module is for generating a 36-D HOG feature vector. In addition to hardware realization, a new method called Histograms-of-Oriented-Gradients AdaBoost Long-Feature-Vector (HOG-AdaBoost-LFV) human classifier is proposed to significantly decrease the number of times to compute the HOG without sacrificing detection performance. The experiment results from three static image and four video datasets demonstrate that the proposed SW/HW (software/hardware) co-design system is 13.14 times faster than the pure software computation of Dalal algorithm.
References
[1]
Dalal, N. and Triggs, B. (2005) Histograms of Oriented Gradients for Human Detection. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, 1, 886-893.
[2]
Zhu, Q., Avidan, S., Yeh, M.C. and Cheng, K.T. (2006) Fast Human Detection Using a Cascade of Histograms of Oriented Gradients. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2, 1491-1498.
[3]
Rabia Rauf, A.R., Shahid, S.Z. and Asad, A.S. (2016) Pedestrian Detection Using HOG, LUV and Optical Flow as Features with AdaBoost as Classifier. 6th International Conference on Image Processing Theory Tools and Applications (IPTA), Oulu, 12-15 December 2016, 1-4.
[4]
Gajjar, V., Gurnani, A. and Khandhediya, Y. (2017) Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach. IEEE International Conference on Computer Vision Workshop (ICCVW), Venice, 22-29 October 2017, 2805-2809. https://doi.org/10.1109/ICCVW.2017.330
[5]
Cao, T.P., Elton, D. and Deng, G. (2012) Fast Buffering for FPGA Implementation of Vision-Based Object Recognition Systems. Journal of Real-Time Image Processing, 7, 173-183. https://doi.org/10.1007/s11554-011-0201-1
[6]
Moranduzzo, T. and Melgani, F. (2014) Detecting Cars in UAV Images with a Catalog-Based Approach. IEEE Transactions on Geoscience and Remote Sensing, 52, 6356-6367. https://doi.org/10.1109/TGRS.2013.2296351
[7]
Madhogaria, S., Baggenstoss, P.M., Schikora, M., Koch, W. and Cremers, D. (2015) Car Detection by Fusion of HOG and Causal MRF. IEEE Transactions on Aerospace and Electronic Systems, 51, 575-590.
https://doi.org/10.1109/TAES.2014.120141
[8]
Ilas, M.-E. (2017) New Histogram Computation Adapted for FPGA Implementation of HOG Algorithm: For Car Detection Applications. 9th Computer Science and Electronic Engineering (CEEC), Colchester, 27-29 September 2017, 77-82.
https://doi.org/10.1109/CEEC.2017.8101603
[9]
Anirudh, B.K., Vivek, V., Abhishek, R.K. and Sumam, D.S. (2017) Accelerating Real-Time Computer Vision Applications Using HW/SW Co-Design. International Conference on Computer, Communications and Electronics (Comptelix), Jaipur, 1-2 July 2017, 458-463. https://doi.org/10.1109/COMPTELIX.2017.8004013
[10]
Bartosz, M., Tomasz, K. and Marek, G. (2017) Embedded Vision System for Pedestrian Detection Based on HOG+SVM and Use of Motion Information Implemented in Zynq Heterogeneous Device. Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA), Poznan, 20-22 September 2017, 406-411.
[11]
Yu, Z., Yang, S., Sillitoe, I. and Buckley, K. (2017) Towards a Scalable Hardware/Software Co-Design Platform for Real-Time Pedestrian Tracking Based on a ZYNQ-7000 Device. IEEE International Conference on Consumer Electronics-Asia (ICCE-Asia), Bengaluru, 5-7 October 2017, 127-132.
https://doi.org/10.1109/ICCE-ASIA.2017.8307853
[12]
Kadota, R., Sugano, H., Hiromoto, M., Ochi, H., Miyamoto, R. and Nakamura, Y. (2009) Hardware Architecture for HOG Feature Extraction. Fifth International Conference on Intelligent Information Hiding and Multimedia Signal Processing, Kyoto, 12-14 September 2009, 1330-1333.
https://doi.org/10.1109/IIH-MSP.2009.216
[13]
Hemmati, M., Biglari-Abhari, M., Berber, S. and Niar, S. (2014) HOG Feature Extractor Hardware Accelerator for Real-Time Pedestrian Detection. Euromicro Conference on Digital System Design (DSD), Verona, 27-29 August 2014, 543-550.
https://doi.org/10.1109/DSD.2014.60
[14]
Niar, M., Miyajima, T. and Amano, H. (2015) Data Reduction and Parallelization for Human Detection System. Workshop on Synthesis and System Integration of Mixed Information Technologies, Yilan, 16-17 March 2015, 134-139.
[15]
Bauer, S., Brunsmann, U. and Schlotterbeck-Macht, S. (2009) FPGA Implementation of a HOG-Based Pedestrian Recognition System. MPC Workshop, Karlsruhe, July 2009, 49-58.
[16]
Bauer, S., Kohler, S., Doll, K. and Brunsmann, U. (2010) FPGA-GPU Architecture for Kernel SVM Pedestrian Detection. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, San Francisco, CA, 13-18 June 2010, 61-68. https://doi.org/10.1109/CVPRW.2010.5543772
[17]
Chen, P.Y., Huang, C.C., Lien, C.Y. and Tsai, Y.H. (2014) An Efficient Hardware Implementation of HOG Feature Extraction for Human Detection. IEEE Transactions on Intelligent Transportation Systems, 15, 656-662.
https://doi.org/10.1109/TITS.2013.2284666
[18]
Suleiman, A. and Sze, V. (2014) Energy-Efficient HOG-Based Object Detection at 1080HD 60 fps with Multi-Scale Support. IEEE Workshop on Signal Processing Systems (SiPS), Belfast, UK, 20-22 October 2014, 1-6.
https://doi.org/10.1109/SiPS.2014.6986096
[19]
Mizuno, K., Terachi, Y., Takagi, K., Izumi, S., Kawaguchi, H. and Yoshimoto, M. (2012) Architectural Study of HOG Feature Extraction Processor for Real-Time Object Detection. IEEE Workshop on Signal Processing Systems (SiPS), Quebec City, 17-19 October 2012, 197-202. https://doi.org/10.1109/SiPS.2012.57
[20]
Takagi, K., Mizuno, K., Izumi, S., Kawaguchi, H. and Yoshimoto, M. (2013) A Sub-100-Milliwatt Dual-Core HOG Accelerator VLSI for Real-Time Multiple Object Detection. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Vancouver, BC, 26-31 May 2013, 2533-2537.
https://doi.org/10.1109/ICASSP.2013.6638112
Marín, J., Vázquez, D., Gerónimo, D. and López, A.M. (2010) Learning Appearance in Virtual Scenarios for Pedestrian Detection. IEEE International Conference on Computer Vision and Pattern Recognition, San Francisco, 13-18 June 2010, 137-144.
[23]
INRIA Person Dataset. http://pascal.inrialpes.fr/data/human
