交通运输系统工程与信息 ›› 2016, Vol. 16 ›› Issue (4): 73-78.

• 智能交通系统与信息技术 • 上一篇    下一篇

基于红外图像和普通图像对比的 高速公路可视度分析

杜豫川*1,张晓明1,刘成龙1,刘翔2   

  1. 1. 上海同济大学道路与交通工程教育部重点实验室,上海201804; 2. 江西赣粤高速公路股份有限公司,南昌330029
  • 收稿日期:2015-12-22 修回日期:2016-03-29 出版日期:2016-08-25 发布日期:2016-08-26
  • 作者简介:杜豫川(1976-),男,四川成都人,教授,博士.
  • 基金资助:

    上海市科学技术委员会研究资助/ Shanghai Science and Technology Committee(14DZ1207204).

Visibility Analysis for Freeway Based on Comparison of Ordinary and Infrared Images

DU Yu-chuan1, ZHANG Xiao-ming1, LIU Cheng-long1, LIU Xiang2   

  1. 1. Key Laboratory of Road and Traffic Engineering of the Ministry of Education, Tongji University, Shanghai 201804, China; 2. Jiangxi Ganyue Expressway Corporation Ltd, Nanchang 330029, China
  • Received:2015-12-22 Revised:2016-03-29 Online:2016-08-25 Published:2016-08-26

摘要:

高速公路可视距离的实时检测对于低能见度预警保证行车安全具有重要意 义,本文旨在建立一种专门应用于高速公路的可视距离检测模型.由此提出可视度概念, 并创新性地通过普通图像和红外图像的像素点的比值来反映实际可视距离,进一步建立 可视度推算模型.同时为减小温度对红外图像影响,提出并推导了温度修正系数;为减小 仪器成像质量差异引入了相机修正系数,有效提高了模型的计算精度.根据离摄像机位置 不同距离范围内像素点的累计百分比,建立可视度查阅表.现场试验结果表明:本文提出 的可视度分析模型预测结果相对误差小于11%,该可视度模型能够在低能见度环境下准 确判断可视距离,可以有效提高高速公路驾驶安全.

关键词: 交通工程, 可视度, 图像对比, 高速公路, 低能见度, 红外图像

Abstract:

The real- time monitoring of the low visibility of the freeway is very important for the traffic safety. This paper is aimed at establishing a visual distance monitoring model which is specially applied to freeway. The visual degree is proposed, and the ratio of the pixel points of the common image and infrared image is presented, which can reflect the actual visual distance, then the paper establishes the visibility model. At the same time, the temperature correction factor is proposed for reducing the impact of temperature on the infrared image, and the correction coefficient of the camera is proposed according to the different imaging quality. Then, several areas are got by dividing the infrared image according range, and according to the distance from the camera, the visual inspection table is established through cumulative percentage of pixels for different areas. Results show that the model has a certain degree of reliability and good accuracy of less than 11%, the visibility model can be used to well determine the visual distance in the low visibility environment, which can effectively improve freeway driving safety.

Key words: traffic engineering, visual level, image comparing, freeway, low visibility, infrared image

中图分类号: