支持行人检测的智能车载监控终端
|
摘要
大量交通事故的分析表明行人是交通的弱势参与者。为了达到保护行人安全的目的,需要提高车辆的智能水平,通过具备智能监控能力的车载终端来协助发现处于事故危险中的行人,及时进行预警。针对车载智能监控的需要,基于聚合积分通道模型,设计了具有行人检测能力的智能移动视频监控终端。面向TI DM365芯片组与TVP5158组成的多路视频处理平台,设计了基于聚合积分通道模型的行人检测算法,并且进行了典型场景实验。实验结果表明,该终端可以通过移动网络实现远程视频监控,并且有效实现行人检测,提高了车辆智能水平与交通安全性。
Analysis of a large number of traffic accidents showed that pedestrians are the vulnerable participants in traffic. In order to protect the safety of pedestrians, it is necessary to improve the intelligent level of vehicles, and assist the detection of pedestrians in danger of accidents through on-board terminals equipped with intelligent monitoring capabilities, so as to timely give early warning. The intelligent mobile video monitoring terminal with pedestrian detection capability has been designed based on the aggregated integral channel model to meet the needs of vehicle-mounted intelligent monitoring. A pedestrian detection algorithm based on the polymer integral channel model has been designed for multichannel video processing platform composed of TI DM365 chipset and TVP5158, and typical scene experiments were conducted. Experimental results showed that the terminal can implement remote video monitoring via mobile network, and effectively complete pedestrian detection, which improves the level of vehicle intelligence and traffic safety.
Analysis of a large number of traffic accidents showed that pedestrians are the vulnerable participants in traffic. In order to protect the safety of pedestrians, it is necessary to improve the intelligent level of vehicles, and assist the detection of pedestrians in danger of accidents through on-board terminals equipped with intelligent monitoring capabilities, so as to timely give early warning. The intelligent mobile video monitoring terminal with pedestrian detection capability has been designed based on the aggregated integral channel model to meet the needs of vehicle-mounted intelligent monitoring. A pedestrian detection algorithm based on the polymer integral channel model has been designed for multichannel video processing platform composed of TI DM365 chipset and TVP5158, and typical scene experiments were conducted. Experimental results showed that the terminal can implement remote video monitoring via mobile network, and effectively complete pedestrian detection, which improves the level of vehicle intelligence and traffic safety.
引文
[1] 齐美彬,李佶,蒋建国,等.改进特征与GPU加速的行人检测[J].中国图象图形学报,2018,23(8):1171-1180.
[2] 常欢,谢永斌.多模车载视频监控系统中负载均衡算法设计[J].电视技术,2018,42(1):90-94.
[3] 陈飞玲,陈湘军,郁建桥,等.移动视频监控系统设计[J].电子测量技术,2014,37(4):103-107.
[4] 钱基德,陈斌,钱基业,等.基于感兴趣区域模型的车道线快速检测算法[J].电子科技大学学报,2018,47(3):356-361.
[5] LLORCA D F, SOTELO M A, HELLíN A M, et al. Stereo regions-of-interest selection for pedestrian protection: A survey [J]. Transportation Research Part C Emerging Technologies, 2012, 25(4):226-237.
[6] BENENSON R, OMRAN M, HOSANG J, et al. Ten years of pedestrian detection, what have we learned? [C]. European Conference on Computer Vision, IEEE, 2014.
[7] LAI C Q, TEOH S S. A review on pedestrian detection techniques based on Histogram of Oriented gradient feature[C]. Research and Development, IEEE, 2014:1-6.
[8] MAJI S, BERG A C, MALIK J. Efficient classification for additive kernel SVMs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013,35(1):66-77.
[9] KELLER C G, ENZWEILER M, ROHRBACH M, et al. The Benefits of dense stereo for pedestrian detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2011, 12(4):1096-1106.
[10] 陈双玉,王海晖,吴云韬,等.基于双目立体视觉和SVM算法行人检测方法[J].华中科技大学学报(自然科学版),2015,43(S1):140-143.
[11] 杨峰,陈建春.基于达芬奇技术的音视频压缩传输系统[J].电子科技,2017,30(8):96-98,101.
[12] 赵太飞,孙孝彬,娄俊鹏.基于H.264的多传输模式IP网络视频监控系统[J].计算机系统应用,2015,24(11):81-87.
[13] ZHAO L, THORPE C E. Stereo-and neural network-based pedestrian detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2000, 1(3):148-154.
[14] ESS A, LEIBE B, SCHINDLER K, et al. A Mobile Vision System for Robust Multi-Person Tracking[C]. 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2008), USA, IEEE, 2008.
[15] DOLLAR P, APPEL R, BELONGIE S, et al. Fast feature pyramids for object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(8):1532-1545.
[16] ZHANG C, VIOLA P A. Multiple-instance pruning for learning efficient cascade detectors[C]. Advances in Neural Information Processing Systems 20, Proceedings of the Twenty-First Annual Conference on Neural Information Processing Systems, Vancouver, British Columbia, Canada, Curran Associates Inc. 2007.
[2] 常欢,谢永斌.多模车载视频监控系统中负载均衡算法设计[J].电视技术,2018,42(1):90-94.
[3] 陈飞玲,陈湘军,郁建桥,等.移动视频监控系统设计[J].电子测量技术,2014,37(4):103-107.
[4] 钱基德,陈斌,钱基业,等.基于感兴趣区域模型的车道线快速检测算法[J].电子科技大学学报,2018,47(3):356-361.
[5] LLORCA D F, SOTELO M A, HELLíN A M, et al. Stereo regions-of-interest selection for pedestrian protection: A survey [J]. Transportation Research Part C Emerging Technologies, 2012, 25(4):226-237.
[6] BENENSON R, OMRAN M, HOSANG J, et al. Ten years of pedestrian detection, what have we learned? [C]. European Conference on Computer Vision, IEEE, 2014.
[7] LAI C Q, TEOH S S. A review on pedestrian detection techniques based on Histogram of Oriented gradient feature[C]. Research and Development, IEEE, 2014:1-6.
[8] MAJI S, BERG A C, MALIK J. Efficient classification for additive kernel SVMs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013,35(1):66-77.
[9] KELLER C G, ENZWEILER M, ROHRBACH M, et al. The Benefits of dense stereo for pedestrian detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2011, 12(4):1096-1106.
[10] 陈双玉,王海晖,吴云韬,等.基于双目立体视觉和SVM算法行人检测方法[J].华中科技大学学报(自然科学版),2015,43(S1):140-143.
[11] 杨峰,陈建春.基于达芬奇技术的音视频压缩传输系统[J].电子科技,2017,30(8):96-98,101.
[12] 赵太飞,孙孝彬,娄俊鹏.基于H.264的多传输模式IP网络视频监控系统[J].计算机系统应用,2015,24(11):81-87.
[13] ZHAO L, THORPE C E. Stereo-and neural network-based pedestrian detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2000, 1(3):148-154.
[14] ESS A, LEIBE B, SCHINDLER K, et al. A Mobile Vision System for Robust Multi-Person Tracking[C]. 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2008), USA, IEEE, 2008.
[15] DOLLAR P, APPEL R, BELONGIE S, et al. Fast feature pyramids for object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(8):1532-1545.
[16] ZHANG C, VIOLA P A. Multiple-instance pruning for learning efficient cascade detectors[C]. Advances in Neural Information Processing Systems 20, Proceedings of the Twenty-First Annual Conference on Neural Information Processing Systems, Vancouver, British Columbia, Canada, Curran Associates Inc. 2007.