基于交通视频分析的自适应摄像控制系统

张洪斌; 黄山; 印月

doi:10.3969/j.issn.0258-2724.2018.03.028

基于交通视频分析的自适应摄像控制系统

doi: 10.3969/j.issn.0258-2724.2018.03.028

张洪斌^1,,
黄山^1,2,
印月²

基金项目:

科技部科技型中小企业技术创新基金资助项目 09C26215102248

四川省重点技术创新项目 07XM020

四川省重点技术创新项目 2014CG053

详细信息

作者简介:
张洪斌(1982-), 男, 博士研究生, 研究方向为计算机视觉在智能交通中的应用, E-mail:zhb593@163.com

中图分类号: TP391.4
计量
- 文章访问数: 360
- HTML全文浏览量: 166
- PDF下载量: 86
- 被引次数: 0
出版历程
- 收稿日期: 2017-12-21
- 刊出日期: 2018-06-01

Adaptive Imaging Control System Based on Traffic Video Analysis

摘要

摘要: 为在高动态光照和复杂路况环境中实现准确的视频检测与数据采集，构建了一套有效的自适应交通摄像控制系统.首先，针对全天候大场景综合检测需求设计高清交通摄像系统，并通过系统辨识确定了核心摄像参数的控制特性；接着，基于车辆号牌和交通场景特点，提出了以车牌亮度中点值和标线区块中点值作为反馈控制指标的视频质量分析算法；最后，使用自适应控制架构综合底层图像质量反馈信息和高层视觉分析检测结果，实现了光照模式的感知适应和控制状态的自主切换.实验与应用结果表明：该系统控制过程快速稳定，能够自动适应不同光照环境，兼顾高清晰车牌识别和大视野交通监控需求，保持良好的全天成像效果，实现了平均97.0%的车流准确度和96.3%的车牌识别率.
- 成像系统 /
- 图像分析 /
- 自适应控制 /
- 车辆检测 /
- 车牌识别 /
- 智能交通系统
Abstract: To ensure accurate video detection and data collection under high dynamic illumination and complicated road conditions, an adaptive traffic imaging control system was developed. A high-definition traffic imaging system was designed to meet the needs of full-time wide-field comprehensive detection. The control characteristics of the imaging parameters were acquired through system identification. Based on license-plate properties and traffic scenes, a video-analysis algorithm computing mid-values of license-plate images and road-marking blocks as feedback control variables, was proposed. In an adaptive control framework combining low-level image-quality feedback and high-level visual-detection results, autonomous illumination-mode adaption and control-state switching were realized. The experimental and application results show that the system control process is fast and stable. It can adapt to different illumination conditions, balance the requirements of high-definition license-plate recognition and wide-field traffic surveillance, maintain good all-day imaging effects, and achieve 97.0% traffic-flow accuracy and a 96.3% license-plate recognition rate.
- imaging systems /
- image analysis /
- adaptive control systems /
- vehicle detection /
- license plate recognition /
- intelligent transport system

HTML全文

图 1 高清交通摄像系统

Figure 1. High-definition traffic-imaging system

下载: 全尺寸图片幻灯片

图 2 摄像系统延时统计

Figure 2. Imaging-system time-lag statistical graph

下载: 全尺寸图片幻灯片

图 3 摄像参数控制特性

Figure 3. Control characteristics of imaging parameters

下载: 全尺寸图片幻灯片

图 4 计算车牌亮度中点值

Figure 4. Computation of license plate mid-value

下载: 全尺寸图片幻灯片

图 5 车牌亮度中点值的分区

Figure 5. Partition of license plate mid-value range

下载: 全尺寸图片幻灯片

图 6 交通场景中的标线区块

Figure 6. Road-marking blocks of a traffic scene

下载: 全尺寸图片幻灯片

图 7 标线区块的可见性判别

Figure 7. Visibility detection of road-marking blocks

下载: 全尺寸图片幻灯片

图 8 自适应摄像控制系统

Figure 8. Adaptive imaging control-system graph

下载: 全尺寸图片幻灯片

图 9 光照模式的感知与控制

Figure 9. Detection and control of different illumination modes

下载: 全尺寸图片幻灯片

图 10 控制状态转换图

Figure 10. Control state-transition graph

下载: 全尺寸图片幻灯片

图 11 光照扰动下的摄像控制效果

Figure 11. Imaging control results with illumination changes

下载: 全尺寸图片幻灯片

图 12 逆光成像效果对比

Figure 12. Imaging effects comparison of backlighting

下载: 全尺寸图片幻灯片

图 13 顺光成像效果对比

Figure 13. Imaging effects comparison of front-lighting

下载: 全尺寸图片幻灯片

图 14 摄像控制量24 h曲线

Figure 14. Twenty-four-hour curves of imaging parameters

下载: 全尺寸图片幻灯片

表 1 特殊光照检测结果

Table 1. Detection results under special illumination conditions

%
摄像控制算法	车牌识别率		场景完整率
摄像控制算法	逆光	顺光	逆光	顺光
算法1	54.8	33.2	96.4	98.0
算法2	94.4	93.2	55.6	86.2
本方法	96.6	96.0	97.8	98.8

下载: 导出CSV

表 2 典型时段成像检测效果

Table 2. Imaging and detection results at typical moments

观测统计项	黎明 7:00—7:30	上午 10:00—10:30	正午 12:00—12:30	下午 15:30—16:00	黄昏 18:30—19:00	子夜 23:30—24:00
车牌图像
路面 T形区块
场景图像
实际过车数/辆	352	1011	816	937	1507	276
车流准确率/%	97.2	97.5	99.1	98.3	94.8	96.7
车牌识别率/%	95.1	98.1	97.2	98.4	94.2	92.1

下载: 导出CSV

参考文献(20)

JIM B, GUSTAVE C, DOMINIE G, et al. ITS strategic plan 2015-2019[R]. Washington D. C.: U.S. Department of Transportation, 2014.

YU C C, CHENG H Y, JIAN Y F. Raindrop-tampered scene detection and traffic flow estimation for nighttime traffic surveillance[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(3):1518-1527. doi: 10.1109/TITS.2014.2365033

ZANG Di, CHAI Zhenliang, ZHANG Junqi, et al. Vehicle license plate recognition using visual attention model and deep learning[J]. Journal of Electronic Imaging, 2015, 24(3):1-10. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=dc5bf6d67000dfdacf1cb32d25e322ce

REN Jianqiang, CHEN Yangzhou, XIN Le, et al. Detecting and positioning of traffic incidents via video-based analysis of traffic states in a road segment[J]. IET Intelligent Transport Systems, 2016, 10(6):428-437. doi: 10.1049/iet-its.2015.0022

KLUBSUWAN K, KOODTALANG W, MUNGSING S. Traffic violation detection using multiple trajectories evaluation of vehicles[C]//Proceedings of International Conference on Intelligent Systems, Modelling and Simulation. [S. l.]: IEEE Computer Society, 2013: 220-224.

ACUNZO D, ZHU Y, XIE B, et al. Context-adaptive approach for vehicle detection under varying lighting conditions[C]//Intelligent Transportation Systems Conference, 2007, IEEE. [S. l.]: IEEE, 2007: 654-660.

O'MALLEY R, JONES E, GLAVIN M. Rear-lamp vehicle detection and tracking in low-exposure color video for night conditions[J]. IEEE Transactions on Intelligent Transportation Systems, 2010, 11(2):453-462. doi: 10.1109/TITS.2010.2045375

LI T, SONG Y, MEI T. An auto exposure control algorithm based on lane recognition for on-board camera[C]//Proceedings of IEEE Intelligent Vehicles Symposium. [S. l.]: IEEE, 2015: 851-856.

黄山.车牌识别技术的研究和实现[D].成都: 四川大学, 2005.

吴洪森, 李志能. "电子警察"视频检测系统的分析与设计[J].浙江大学学报:工学版, 2005, 39(10):1517-1519. http://d.old.wanfangdata.com.cn/Periodical/zjdxxb-gx200510012

WU hongsen, LI Zhineng. Analysis and design of auto-monitoring and recording system for electronic police[J]. Journal of Zhejiang University:Engineering Science, 2005, 39(10):1517-1519. http://d.old.wanfangdata.com.cn/Periodical/zjdxxb-gx200510012

潘薇.嵌入式车牌识别系统关键技术研究[D].成都: 四川大学, 2010.

RAGHAVAN A, LIU J, SAHA B, et al. Reference image-independent fault detection in transportation camera systems for nighttime scenes[C]//Proceedings of IEEE Conference on Intelligent Transportation Systems. [S. l.]: IEEE, 2012: 963-968.

TORRES J, MENÉDE J M. Optimal camera exposure for video surveillance systems by predictive control of shutter speed, aperture, and gain[C]//Real-Time Image and Video Processing. [S. l.]: SPIE, 2015: 1-14.

黄凯奇, 陈晓棠, 康运锋, 等.智能视频监控技术综述[J].计算机学报, 2015, 38(6):1093-1118. http://d.old.wanfangdata.com.cn/Periodical/jsjxb201506001

公安部道路交通管理标准化技术委员会. GA/T 496-2014闯红灯自动记录系统通用技术条件[S].北京: 中华人民共和国公安部, 2014.

OTSU N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems Man & Cybernetics, 1979, 9(1):62-66. http://jamia.oxfordjournals.org/lookup/external-ref?access_num=10.1109/TSMC.1979.4310076&link_type=DOI

张洪斌, 黄山.面向城市路口的高清晰智能监控系统研究[J].四川大学学报:工程科学版, 2012, 44(增刊1):224-228. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK201201519114

ZHANG Hongbin, HUANG Shan. Research on the high definition intelligent surveillance system for urban intersection[J]. Journal of Sichuan University:Engineering Science Edition, 2012, 44(Sup.1):224-228. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK201201519114

MAMMERI A, BOUKERCHE A, TANG Z. A real-time lane marking localization, tracking and communication system[J]. Computer Communications, 2015, 73(2):229-233. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=1cda0e292c1f21d1e874d4255a45d480

MATAS J, KITTLER J, GALAMBOS C. Robust detection of lines using the progressive probabilistic Hough transform[J]. Computer Vision and Image Understanding, 2000, 78(1):119-137. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=21e5c3051d3df8c83b84714e97113bb6

公安部道路交通管理标准化技术委员会. GA/T 833-2016机动车号牌图像自动识别技术规范[S].北京: 中华人民共和国公安部, 2016.

施引文献

附加材料(0)

访问统计