The Realization Principle and Method of Qingdao Road Gate and Qingdao License Plate Recognition Technology

The Realization Principle and Method of license plate recognition Technology



Image acquisition: Real-time, uninterrupted recording and acquisition of vehicle violation or passing through the jack by high-definition camera capturing host. Pre-processing: Picture quality is the key factor affecting vehicle recognition rate. Therefore, it is necessary to capture high-definition camera.

The original image collected by the host computer is processed by noise filtering, automatic white balance, automatic exposure, gamma correction, edge enhancement and contrast adjustment.



License Plate Location: The accuracy of license plate location directly determines the character segmentation and recognition effect, which is an important factor affecting the whole license plate recognition rate. The core of the algorithm is texture feature analysis and location algorithm. After image preprocessing, the gray image is obtained.

In the image, row scan is used to determine the candidate area containing the license plate line segment in the column direction, the initial row coordinates and the height of the area are determined, and then the column scan is used to determine the column coordinates and the width of the area, so as to determine a license plate area. This algorithm can locate all license plates in the image.



Character segmentation: After locating the license plate area in the image, the character area is further accurately located by gray-scale, gray-scale stretching, binarization and edge processing. Then, according to the character size characteristics, a dynamic template method is proposed for character segmentation.

The character size is normalized.



Character Recognition: Scale and extract the features of the segmented characters to obtain the expression of specific characters, and then match and discriminate with the standard character expression in the character database template through classification discriminant function and classification rules.

To recognize the input character image.



Output of result: The result of license plate recognition is output in text format.



Vehicle license plate recognition is based on image segmentation and image recognition theory.



The image containing vehicle number plate is analyzed and processed.

Thus, the position of the license plate in the image is determined, and the text characters are further extracted and recognized.



Recognition steps are summarized as: license plate location, license plate extraction, character recognition. Three steps of recognition work complement each other, their respective efficiency is higher, the overall recognition rate will be improved.



The speed of recognition depends on character recognition. At present, the main application technology of character recognition is comparative recognition sample library.



All the characters are set up in the sample library, and the character is judged by comparing the sample library after character extraction.



In the process of recognition, intermediate result values such as credibility and inclination will be generated, and the other is character recognition technology based on character structure knowledge.



It is more effective to improve the recognition rate and accuracy, and has strong adaptability.



There are two main ways to realize the license plate recognition system: one is the recognition of static images and pictures.



The other is real-time identification of dynamic video streams.



The recognition efficiency of static image recognition technology is largely limited by the quality of image capturing. At present, the average recognition speed of products in the market is 200 milliseconds, while the recognition technology of dynamic video stream has strong adaptability and fast recognition speed.

It realizes the recognition of each frame of video image, increases the number of recognition and comparison, and chooses the best license plate number.





The key is that it is less affected by the quality of single frame image. At present, the time of product recognition in the market is 10 milliseconds.



What is the principle of license plate recognition? The license plate recognition parking management system automatically recognizes and converts the image of the license plate number taken by the camera at the entrance into a digital signal. The advantage of license plate recognition is that it can match the card with the car.

To improve the management, the corresponding advantage of the card and the car is that the long-Rent card whisker is used in conjunction with the car to eliminate more than one card.



Vehicle use loopholes, improve the efficiency of property management; at the same time, automatic comparison of vehicles in and out to prevent theft incidents. The upgraded camera system can collect clearer pictures for archival preservation and provide strong evidence for some disputes.

It is convenient for managers to compare when the vehicle exits, and greatly enhances the security of the system.



Automatic license plate recognition technology is a pattern recognition technology which uses dynamic video or static image of vehicle to recognize license plate number and color automatically. Through image acquisition and processing, the automatic license plate recognition function is completed.

It can automatically extract license plate image from an image, automatically segment characters, and then recognize the characters. Its hardware basis generally includes trigger equipment (monitoring whether the vehicle enters the field of vision), camera equipment, lighting equipment, image acquisition equipment, etc.

The software core of license plate recognition processor (such as computer) includes license plate location algorithm, license plate character segmentation algorithm and optical character recognition algorithm. Some license plate recognition systems also have the ability to judge vehicle entry through video images.

The function of the field is called video vehicle detection. A complete license plate recognition system should include vehicle detection, image acquisition, license plate recognition and other parts (as shown in Figure 1). When the vehicle detection part detects the arrival of the vehicle, it triggers the image acquisition unit.

Acquisition of current video images. The license plate recognition unit processes the image, locates the position of the license plate, and then divides the characters in the license plate for recognition, and then composes the output of the license plate number.



1. Vehicle detection can adopt buried coil detection, infrared detection, radar detection, video detection and other methods. Video detection can avoid damage to the road surface, no additional external detection equipment, no need to correct the trigger position.

It saves money and is more suitable for mobile and portable applications.



A license plate recognition system with video vehicle detection function is presented. Firstly, a frame (field) signal in the video signal is plotted.