Vehicle License Plate RecogniTIon (VLPR) refers to a technology that can detect vehicles on a monitored road surface and automatically extract vehicle license plate information (including Chinese characters, English letters, Arabic numerals, and number plates) for processing. License plate recognition is one of the important components of modern intelligent transportation systems, and it is widely used. Based on digital image processing, pattern recognition, computer vision and other technologies, it analyzes the vehicle image or video sequence captured by the camera to obtain the unique license plate number of each car, thus completing the identification process. Through some subsequent processing methods, it can realize the functions of parking lot charge management, traffic flow control index measurement, vehicle positioning, automobile anti-theft, highway over-speed automatic supervision, red light electronic police, highway toll station and so on. It is of practical significance to maintain traffic safety and urban security, prevent traffic jams, and realize automatic traffic management.
The license plate recognition technology is combined with the electronic non-stop toll collection system (ETC) to identify the vehicle. When the passing vehicle passes through the crossing, there is no need to stop, that is, the vehicle identity can be automatically recognized and automatically charged. In the management of the yard, in order to improve the traffic efficiency of the entry and exit vehicles, the license plate recognition is for vehicles that do not need to collect parking fees (such as monthly trucks, free internal vehicles), and build an unattended fast-track, free of access and non-stop access experience. Change the management mode of access to the parking lot.
License plate recognition principle explains license plate recognition process
1, image acquisition
According to different vehicle detection methods, image acquisition is generally divided into two types. One is image acquisition in static mode. The vehicle triggers a sense coil, infrared or radar to give the camera a trigger signal, and the camera receives the trigger signal. After taking an image, the method has the advantages of high trigger rate and stable performance. The disadvantage is that the coil is required to be cut on the ground, and the construction amount is large. The other is image acquisition in video mode, and no external trigger signal is required. The video stream image will be recorded in real time. The advantage of this method is that it is convenient to construct, does not need to cut the ground to lay the coil, and does not need to install parts such as the vehicle inspection device, but its disadvantages are also very significant. Due to the limitation of the algorithm, the scheme triggers. The rate and recognition rate are lower than the external trigger. After the strict algorithm optimization in Beijing Yibo era, the recognition rate and stability of these two image acquisition modes are among the highest in the industry.
2, pretreatment
Since the image quality is easily affected by factors such as light, weather, camera position, etc., it is necessary to pre-process the camera and image before identifying the license plate to ensure the clearest image of the license plate. Generally, based on the analysis of the scene environment and the images already captured, the camera realizes automatic exposure processing, automatic white balance processing, automatic backlight processing, automatic over-explosion processing, etc., and performs noise filtering and contrast enhancement on the image. Image scaling and other processing. Denoising methods include mean filtering, median filtering and Gaussian filtering; contrast enhancement methods include contrast linear stretching, histogram equalization and homomorphic filters; the main methods of image scaling are nearest neighbor interpolation and bilinear interpolation. Method and cubic convolution interpolation.
3, license plate positioning
Accurate detection of the license plate area from the entire image is an important step in the license plate recognition process. If the positioning fails or the positioning is incomplete, it will directly lead to the final recognition failure. Due to the complex image background and the unclear license plate positioning, it is easy to treat the noise such as fences and billboards as license plates, so how to eliminate these fake license plates is also a difficult point for license plate positioning. In order to improve the accuracy of positioning and improve the recognition speed, the general license plate recognition system will design an external interface, allowing users to set different identification areas according to the site environment. The easy parking license plate recognition system can be used to set the identification area for the scene of some complex backgrounds (such as green belts and manhole covers).
4, license plate correction
Due to the influence of shooting angle, lens and other factors, the license plate in the image has horizontal tilt, vertical tilt or trapezoidal distortion, which brings difficulties to the subsequent recognition processing. If the license plate correction process is performed after the license plate is positioned, this is advantageous for removing noise such as the license plate frame and is more advantageous for character recognition. At present, the commonly used correction methods are: Hough transform method, which calculates the tilt angle by detecting the straight line of the license plate up and down and the left and right borders; the rotary projection method, by vertically projecting the image on the horizontal axis according to different angles, the sum of the points whose projection value is 0 The maximum angle is the vertical tilt angle, and the horizontal angle is calculated similarly. The principal component analysis method has a fixed matching function according to the color of the license plate background and the character boundary, and finds the principal component direction of the color versus the feature point. The horizontal tilt angle of the license plate; the minimum variance method, the closed expression of the vertical tilt angle is derived according to the minimum coordinate deviation of the projected point of the character in the vertical direction, thereby determining the vertical tilt angle; the perspective transformation, using the four vertices of the detected license plate The distortion correction of the license plate is realized after the correlation matrix is ​​transformed. The Yibo era still has a high recognition rate for large-angle license plate recognition.
5, character segmentation
After positioning the license plate area, since there is no information about the total number of characters in the license plate, the positional relationship between the characters, the width and height of each character, etc., in order to ensure the correct matching of the license plate type and the correct character recognition, character segmentation is essential. One step. The main idea of ​​character segmentation is based on the binarization result of the license plate or the edge extraction result, using the structural features of characters, the similarity between characters, the interval between characters, etc., on the one hand, extracting individual characters separately, including adhesion and The processing of special cases such as broken characters; on the other hand, the characters of wide and high similarity are classified into one class to remove the license plate border and some small noise. The commonly used algorithms are: connected domain analysis, projection analysis, character clustering and template matching. Fuzzy license plates caused by dirty license plates and uneven illumination are still the challenges faced by the character segmentation algorithm, and better algorithms are needed to solve the above problems.
6, character recognition
The grayscale image of the segmented character is normalized, the feature is extracted, and then machine learning or matching with the character database template is performed, and finally the result with the highest matching degree is selected as the recognition result. Currently popular character recognition algorithms are: template matching method, artificial neural network method, support vector machine method and Adaboost classification method. The advantage of the template matching method is that the recognition speed is fast and the method is simple. The disadvantage is that it has some difficulties in dealing with the fracture and fouling; the artificial neural network method has strong learning ability, strong adaptability, strong classification ability but time-consuming; support vector The machine method has better recognition ability for the test samples that have not been seen and requires less training samples. The Adaboost classification method can focus on the more important training data, and the recognition speed is fast and the real-time performance is high. China's license plate consists of three characters: Chinese characters, English letters and Arabic numerals, and has a uniform style, which is also a convenient part of the recognition process. However, since the license plate is easily affected by the external environment and there are blurring, breakage, and staining characters, how to improve the recognition rate of such characters and confusing characters is also one of the difficulties in character recognition. The confusing characters include: 0 and D, 0 and Q, 2 and Z, 8 and B, 5 and S, 6 and G, 4 and A, and the like.
7, license plate recognition result output
The license plate recognition result is output in text format, including license plate number, license plate color, license plate type, and the like.
If you are a professional or amateur hacker, then you have no such troubles at all. The black entry license recognition system is as simple as a takeaway. You only need to click on the Internet to enter your license plate identification number on the system. And related information, then your car can enter and exit the cell unconditionally. Although this method is a little risky, it is not a hacker who is not caught in a big event. It seems that few hackers have not even found it.
2, deal with the security uncleEven the best machines will have a "strike" time, so every technology product has a plan to deal with strikes, and the license plate recognition product is a remote control equipped with a remote control gate. That is to say, in the hands of the security uncle, there is a remote control that can control the switch. As long as your relationship with him is good enough, it is not very easy to open the door for you every day.
3, the deckThe deck is the most direct method, and it is also the most widely used method. It can be eaten not only in the community but also in places where license plate recognition is available. But this method is the most risky of all methods. It is caught by the police uncle to arrest, fine and detain, so don't use this method unless you have to.
Recommended reading:
License plate recognition error _ license plate recognition can not be solved
Development and Significance of License Plate Recognition Technology_Introduction to the Principles of License Plate Recognition System
Which license plate is recognized? The top ten license plate recognition brands are listed in the national rankings.
Scsi Connector,26P Scsi Connector,Solder Male With Latch Connector,2.16Mm Solder Male Connector
Shenzhen Hongyian Electronics Co., Ltd. , https://www.hongyiancon.com