Distance measurement method of binocular stereoscopic gazing monitoring system

Abstract

The invention discloses a distance measurement method of a binocular stereoscopic gazing monitoring system. The distance measurement method comprises the following steps that firstly, the coordinates of a principal point of an image, the focal length of a left camera and the focal length of a right camera are determined; secondly, the postures of the binocular cameras are adjusted so as to enable the binocular cameras to gaze at a target point; thirdly, a target distance measurement model based on a rotation angle is established; fourthly, the adjusted binocular convergence cameras are calibrated stereoscopically; fifthly, a transformational relation between the rotation angle and relative external parameters is established; sixthly, target distance measurement and coordinate reestablishment are carried out on the basis of the relative external parameters. On the basis of the distance measurement model, after corresponding parameters are introduced, the distances between the target point and the binocular cameras and the space three-dimensional coordinates of the target point can be worked out. According to the distance measurement method of the binocular stereoscopic gazing monitoring system, the measurement accuracy is high, the detectable distance is long, the result is worked out rapidly and accurately, and the requirement for target positioning of video monitoring can be met.

Description

technical field [0001] The invention belongs to the technical field of a binocular stereo camera, and further relates to a ranging method of a binocular stereo gaze monitoring system. technical background [0002] Video surveillance systems have been widely used in many fields of social life, bringing great convenience and safety to people's social life. At present, most of the video monitoring nodes are single-viewpoint plane monitoring systems, which acquire images from one viewpoint. It can't complete the precise positioning of the target, brings some changes to the users of video surveillance, and can't meet the higher requirements of intelligent surveillance. [0003] The binocular stereo vision technology can well make up for the shortcomings of single-view plane monitoring that cannot accurately locate the target. Binocular stereo vision directly simulates the processing method of human eyes to collect stereo images, has the ability to obtain object depth informatio...

AI Technical Summary

Problems solved by technology

[0005] However, the binocular stereo vision system based on the principle of parallax has some shortcomings: First, the current binocular stereo image pair matching is for the entire image information, and stereo matching is a very time-consuming task

For a general video surveillance system, it does not need to know the depth information of the entire scene but only needs to know the depth information of a specific target. Obviously, this method has a lot of information redundancy and time consumption.

Secondly, the implementation steps of the binocular stereo vision system based on the principle of parallax are relatively complicated, and there are many algorithms required and the robustness requirements of each algorithm are high. In particular, the stereo matching algorithm has not yet had a very reliable solution.

Then, for some specific video surveillance situations (such as traffic video surveillance), the required target detection distance is several meters to tens of meters, and the detection distance of this method is very limited under the condition of ensuring the measurement accuracy.

Images