摘 要

随着经济的快速发展，高速公路的应用也变得更加广泛，同时也是物流运输发展必需品。目前，我国的高速公路总里程已达到14万公里。随之而来的问题是如何在高速公路上做到安全快速的运输物流产品，在风霜雨雪等恶劣天气行车发生交通事故的概率很高，特别是遇到雾、沙尘暴等突发情况时，更容易发生一系列追尾事故，追尾事故中的小型车辆驾驶员很容易发生死亡，而且处理车祸中装满货物的物流车辆，需要大量的人力物力的消耗，对物流的配送也造成了巨大的影响，物流的时效性不能保证的话，同样会造成巨大的人员浪费以及经济损失。根据团雾的特点为了人民财产和物流行业的安全着想，需要开发出一套团雾智能感知系统，实时的检测团雾，可以有效的避免高速公路交通事故的发生。

为了解决和减少高速公路的交通安全问题，必须有个仪器或者系统能快速得检测团雾，也就是检测高速公路的能见度，本文的团雾智能感知系统是一个车载的基于图像分析的系统，为了设计这个系统，本文先是简单的介绍了团雾对高速公路交通安全带来的影响，之后用详细的交通事故数据来证明团雾的危害性，之后查询资料获得团雾形成的气象条件，通过资料分析出团雾的特点；又详细的分析了团雾引发的交通事故的详细过程，得出了团雾现象给物流行业带来的不利影响巨大的结论。从上得出研究团雾预警系统的重要性。分析了国内外的团雾研究的发展现状，之后详细的分析了国外研究出来的仪器，着重写出了德国、芬兰、英国、美国仪器的详细数据，紧接着介绍了国内研究发明的5种能见度仪器，并在图表中给出了具有代表性的公司，最后将国内外现有的仪器进行对比，分析仪器之间的优缺点，选择了最适合本文的图像视频分析技术以开发系统。从自然光线的角度出发，详细的分析自然光在运动过程中的各种情况，为了解决这样的影响，选择激光为发射端，带入适当的误差值，得出以 Koschmieder 模型为基础，修正后的能见度公式；之后为了降低时间因素的干扰，分类分析得出白天、黑夜、白天黑夜过渡的算法模型，分别以光强衰变、种子区域生长法、基于图像灰度、基于经纬度的算法进行检测。根据上述的算法模型，实现团雾智能感知系统的设计工作，先是介绍了团雾具体的检测流程，然后详细的介绍了流程，之后分类讨论有无数据条件下的信息传递方式，在利用车联网的技术解决通信问题；紧接着介绍了系统的图像处理功能，在考虑到各种因素的干扰，分别多种方式来完成团雾图像的分析过程，证实了软件的可行性。在设计出软件的前提下，详细的介绍了系统的使用，以及系统后台的处理流程，在信息传递方面，可以根据前方的情况而定，自主选择使用什么样的通讯方式。

关键词：物流安全；高速公路；团雾；智能感知

Intelligent Sensing System of Highway Cluster Fog for Logistics Safety

ABSTRACT

With the rapid development of economy, the application of expressway has become more and more extensive, and it is also a necessity for the development of logistics and transportation. At present, the total mileage of Expressway in China has reached 140,000 kilometers. The following problem is how to transport logistics products safely and quickly on the expressway. The probability of traffic accidents is very high in bad weather such as wind, snow, rain and snow. Especially in the case of sudden fog, dust storms and other emergencies, a series of rear-end accidents are more likely to occur. Small vehicle drivers in rear-end accidents are prone to death, and the handling of car accidents is full of goods. Logistics vehicles need a lot of manpower and material resources consumption, which also has a huge impact on the distribution of logistics. If the timeliness of logistics cannot be guaranteed, it will also cause huge waste of personnel and economic losses. According to the characteristics of cluster fog, for the sake of the safety of people's property and logistics industry, it is necessary to develop an intelligent sensing system of cluster fog, which can detect the cluster fog in real time and effectively avoid the occurrence of highway traffic accidents.

In order to solve and reduce the traffic safety problem of expressway, it is necessary to have an instrument or system to detect the mass fog quickly, that is, to detect the visibility of expressway. The intelligent sensing system of mass fog in this paper is a vehicle-mounted image-based analysis system. In order to design this system, the impact of mass fog on the traffic safety of expressway is briefly introduced in this paper. After that, detailed traffic accident data are used to prove the hazards of the cluster fog. After that, the meteorological conditions of the formation of the cluster fog are obtained by inquiring the data, and the characteristics of the cluster fog are analyzed through the data. The detailed process of the traffic accident caused by the cluster fog is also analyzed in detail, and the conclusion that the adverse impact of the cluster fog phenomenon on the logistics industry is enormous is drawn. From the above, the importance of the early warning system of the study group fog is concluded. This paper analyses the development status of the research on the mass fog at home and abroad, and then analyses the instruments developed abroad in detail. The detailed data of German, Finnish, British and American instruments are emphatically written. Five kinds of visibility instruments invented in China are introduced, and representative companies are given in the chart. Finally, the existing instruments at home and abroad are compared and analyzed. The most suitable image and video analysis technology is chosen to develop the system. From the point of view of natural light, various situations of natural light in the process of movement are analyzed in detail. In order to solve this problem, laser is chosen as the emitter, and appropriate error values are brought in. Based on Koschmieder model, the revised visibility formula is obtained. Then, in order to reduce the interference of time factors, the transition of day, night, day and night is classified and analyzed. The algorithm model is based on intensity decay, seed region growth, image gray level and longitude and latitude. According to the above algorithm model, the design of the intelligent sensing system for fog is realized. Firstly, the specific detection process of fog is introduced, then the flow is introduced in detail. Then, the information transmission mode under the condition of no data is discussed, and the communication problem is solved by using the technology of vehicle networking. Then, the image processing function of the system is introduced, taking into account the interference of various factors. The analysis process of the fog image is completed in different ways, which proves the feasibility of the software. On the premise of designing the software, the use of the system and the processing flow of the system background are introduced in detail. In the aspect of information transmission, according to the situation ahead, we can choose the communication mode independently.

Key words：Logistics Safety；Expressway；Cluster fog；Intelligent perception

目录

1 绪论 - 1 -

1.1 研究背景 - 1 -