Application of Digital Image Processing Technology in the Field of Intelligent Trasportation

电子科技大学 格拉斯哥学院 2017级 刘奕昕，同组同学 屈颖珊
The concept of computer and image processing technology, so-called computer image processing (computer image processing) is to first convert the analog image signal into a digital signal, and then use the advantages of computer processing of digital signals, the digital image is translated, rotated, denoised Methods and techniques for smoothing, enhancing, splitting, compressing, etc., therefore also known as digital image processing. As a result, this technology is very useful in many fields. This essay will mainly introduce its application in Intelligent Transportation System.

Driven by China’s economic development, the number of people owning cars has gradually increased, which puts higher demands on the development of highways in China. Especially in the transportation system, for the efficient management of safe travel and traffic conditions, digital image processing technology has been introduced to enable intelligent traffic management. In order to improve the chaotic traffic conditions, reduce congestion, improve transportation efficiency and improve traffic safety, various high-tech technologies are used to systematically solve road traffic problems, and new research and application fields have emerged - Intelligent Transportation System ITS (Intelligent Transportation System). It is a hot spot in the competition of international transportation technology and information technology. Digital image processing technology can be regarded as a general term for various image processing technologies in a broad sense. It is an emerging technology field that has been developed, developed and matured with the development of computer technology and VLSI since the 1960s. The amount of data processed by digital image processing is very large, so the computer computing power is rapidly increased; the running speed is greatly improved; the price is rapidly decreasing, and the image processing equipment is rapidly transitioning from medium to small computers to personal computers, all of which are digital image processing technologies in the field of ITS. Wide range of applications provide the conditions. Digital image processing technology plays an important role in the field of ITS and has broad application prospects. Vehicle license plate recognition system combining computer vision and pattern recognition technology is an important part of intelligent transportation system. It can be used for traffic flow detection; traffic control and induction; vehicle management in airports, ports and communities; automatic charging without parking; public Parking security and anti-theft management. This article will focus on the digital image processing techniques used in license plate recognition systems.

Applying digital image processing technology to traffic can realize the intelligentization of traffic, greatly facilitating traffic, and effectively avoiding many problems in traffic. In this paper, the application of digital image processing technology in intelligent transportation is analyzed, and the practical application of digital image processing technology in intelligent traffic is integrated. Some examples are as following:

1. Tracking vehicle movements and paths
   In actual traffic, many accidents may occur. In order to effectively understand the events and improve the actual work efficiency, digital image processing technology can be used to track vehicle motion and path. In this respect, the main tracking detection is the actual operating state of the vehicle in the case of parking, accident, congestion, lane change, and the like. Compared with the traditional technical means, this technology has obvious advantages in this respect. The application of this technology does not need to damage the road, which can effectively reduce the difficulties in the installation process, and secondly, it will not damage the service life of the road . The working principle of this technology is to monitor and record the time running status of the vehicle, and the captured video is transmitted to the digital image technology for calculation and analysis.

2. Identify obstacles
   In the actual traffic management operation, the identification of obstacles is also a very important task. If the driver is driving at a high speed on the highway, but encounters a large obstacle at a close distance, then he wants to brake or It is very difficult to hit the steering wheel urgently, and it is easy to have a traffic accident. However, the digital image processing technology can detect the road conditions in time. Once the large or abnormal obstacles are found, the relevant units can be notified in time, and the relevant work units can be repaired and cleaned. In this way, traffic accidents caused by debt can be effectively reduced, and other traffic safety is improved. There are many methods that can be utilized in this respect, and the common ones are image filtering method, template method, and the like. In this respect, the working principle of the digital image processing technology is to monitor the road condition by means of the camera, identify the abnormal and bulky obstacles, and then transmit the monitored alarm signal to the relevant monitoring platform, and the corresponding processing is performed by the monitoring platform.

Just as mentioned, processing technology has extremely wide applications in the field of ITS, but the most direct and most interesting is the license plate recognition system. The license plate recognition system is a computer vision system that uses the license plate as a specific target and automatically recognizes other license information such as license plate number and license plate color through image recognition. It is an important research content of computer vision and pattern recognition technology in the field of intelligent transportation. Modern intelligent transportation has an important meaning. In a complete license plate recognition system, when the license plate image is acquired, the obtained image is identified. License plate recognition generally involves several steps: image preprocessing, license plate location, character segmentation, and character recognition. In the process of recognizing the license plate, the main thing is the processing and recognition of the picture. The process is further embodied by first transmitting the picture collected from the scene to the digital image processing computer, and the collected image is blurred, and There are a lot of background noises, which is not conducive to the extraction of information. Therefore, the next step is to pre-process the image, improve the quality of the image, and then locate the license plate area with some knowledge of mathematics and corresponding technology. Finally, the extracted license plate image is binarized and mathematically filtered, and finally filtered to correct the license plate position picture information. Finally, the license plate image is segmented and recognized for characters, letters, numbers, and the like.

The image pre-processing program performs filtering, boundary enhancement, and the like on the acquired license plate image for subsequent processing. The license plate location module finds the location of the license plate area from the processed image and marks it to segment the license plate area from the complex background. It is the premise of license plate character recognition and the key technology for license plate recognition. Character segmentation is the division of the extracted license plate area into individual characters. The character recognition module is one of the core modules of the license plate recognition system, and the advantages and disadvantages of the algorithm directly affect the system recognition rate. In the process of character recognition, due to the existence of many interference factors in actual use conditions, these put forward high requirements for the recognition algorithm. Through such a process, the identification and processing of vehicle license plate information is completed, so that vehicle violations, traffic accidents, and the like can be effectively monitored and managed.

In conclusion, digital image processing technology has become one of the most important technologies in its field. In the aspect of image processing hardware, although the related hardware technology is developing rapidly, the development of special hardware for image processing should be strengthened. In the aspect of image processing software, in order to solve the problem of ambiguity in detection and recognition, more adaptable algorithms should be studied. In a word, with the rapid development of computer technology, image processing for its will certainly make great progress. In intelligent transportation, the implementation of digital image processing technology has promoted the development and promotion of intelligent transportation to a certain extent, but it is more important to note that in order to realize the long-term, effective and scientific development of intelligent transportation, it is still necessary to continuously improve and innovate the technology.