Vehicle license plate recognition-based multi-lane vehicle distance recognition method and device

Abstract

The invention provides a vehicle license plate recognition-based multi-lane vehicle distance recognition method and a vehicle license plate recognition-based multi-lane vehicle distance recognition device and belongs to the intelligence assisted driving / vehicle-mounted information detecting field. According to the vehicle license plate recognition-based multi-lane vehicle distance recognition method and the vehicle license plate recognition-based multi-lane vehicle distance recognition device, a binocular camera is adopted to acquire the image information of a plurality of lanes ahead in real time; according to a characteristic that the RGB information of pixels on a vertical splicing line of a left image and a right image is consistent, and with the preset coordinate point of the installed device is used in a default manner, a multi-lane panoramic image is obtained. In an image recognition process, a plurality of classic digital image processing algorithms dominated by HOUGH transform are used to calibrate lane lines and license plates; and based on the positions of different license plates and the sizes of pixels in the panoramic image, and according to a similarity principle and a digital imaging principle, multi-lane horizontal and vertical vehicle distances are automatically calculated, and calculation results are displayed on a screen according to the settings of a user. The method and device can be applied to intelligent traffic fields such assisted safety lane changing, active collision avoidance and ACC adaptive cruise.

Description

technical field [0001] The invention relates to the field of intelligent assisted driving / vehicle information detection, in particular to a method and device for acquiring multi-lane images in real time through a binocular camera and identifying horizontal and vertical inter-vehicle distances from the image information. Background technique [0002] As people pay more and more attention to driving safety and the cost of supporting hardware decreases, more and more body sensors are installed on cars to collect various data to provide assisted driving or active safety services. Among them, the ability to perceive the surrounding environment of the vehicle is The most important thing is that with a more comprehensive understanding of the driving environment, accidents caused by drivers' inattention or blind spots can be effectively reduced. While driving a vehicle, the driver's estimation of the distance between vehicles is very important, but it is difficult to guarantee that ...

AI Technical Summary

Problems solved by technology

The former has higher accuracy and is less likely to be misjudged but has a smaller detection range. It is currently mostly used in parking assistance systems, while the latter is rarely used in civilian vehicles due to the high cost of such products on the market and the difficulty of installation.

[0015] 3 As for the in-vehicle image assistance system, the reversing image system is mostly used in the car at present, and only one camera is installed on the rear side of the vehicle body to provide the driver with rear influence for reversing. For normal driving conditions, such as monitoring the front environment Judgment scheme design is almost non-existent

On the other hand, the visual sensors used in automobiles generally only provide real-time images for the driver and do not perform image-based processing. The main reason is that high-performance digital image processing systems are limited by cost and volume and are difficult to use in vehicles. However, some simple microprocessor-based digital image processing systems can hardly meet the real-time requirements, and cannot feed back effective information in time.

Images