Video detection system and data processing device thereof, video detection method

Abstract

The invention relates to a video detection system and a data processing device and a video detection method thereof, wherein the video detection method comprises the following steps: the traffic video flow is acquired, and the acquired video flow is transmitted to the processing device; the processing device extracts recognition ancillary information from each frame of image; and if a car head is detected in the image according to the recognition ancillary information, the follow-up images are detected and a road traffic information parameter is generated. The invention can achieve full video detection to the road traffic information parameter, avoid false detection and miss detection caused by vehicle driving over lines, vehicle shadow, outside light change and the like, and increase the detection efficiency and accuracy. In addition, the invention can also achieve the real-time display of a dynamic image, and the monitoring of road traffic condition.

Description

technical field [0001] The invention relates to a video detection system and method, in particular to a method for full video detection of road traffic information, a video detection system and a data processing device thereof. Background technique [0002] Road traffic information is an indispensable and important data in road planning, design and maintenance management, among which, road traffic information can specifically include various information such as traffic volume, vehicle speed, time occupancy rate, and headway. With the development of the economy, the traffic flow is increasing rapidly, and it is far from enough to rely solely on manual statistics. In this case, the automatic collection system of road traffic information is being applied more and more. [0003] The automatic collection technology of road traffic information includes contour scanning detection technology, magnetic field change detection technology and video detection technology. Video detectio...

AI Technical Summary

Problems solved by technology

There are some deficiencies in the technical solution: 1), the front end needs to install a computer and an image acquisition card for each camera, which greatly increases the number of devices in the system, and the installation and maintenance costs of the devices are relatively high; 2), the computer determines In the process of whether there is a car arriving, only the road background feature image is scanned, which may cause misjudgment due to the presence of some images of other cars in the background feature image, which may cause unnecessary or wrong shots by the camera; 3) , the image processed by the computer is received by the remote monitoring center at the back end, and it is possible to miss key images

[0006] However, in the highway traffic flow detection system, the frame difference method is used to determine whether there is a vehicle on the detected road, and it is difficult to avoid the occurrence of misjudgment

For example, when parts of other cars or other non-vehicle objects enter a predetermined area in the image, it may cause a large change in the pixel value of the area in the previous and subsequent frame images, even exceeding the set threshold

In addition, the detection system obtains less road traffic information, and basically can only roughly estimate the traffic flow, and it is difficult to obtain various other traffic parameters about vehicles in motion, such as vehicle speed, vehicle type, etc. That is to say, all The above detection system cannot realize the monitoring of the vehicle

Images