Target positioning and tracking method and device thereof

Abstract

The invention relates to the field of video monitoring and discloses a target positioning and tracking method and device. The target positioning and tracking method comprises the following steps that depth data are acquired and converted into depth image data, wherein the gray value of the depth image data indicates the distance between each data point in the depth image data and an acquisition device; prospect data of the depth image data are extracted; the three-dimensional coordinates of the prospect data in the acquisition device are calculated, and the three-dimensional coordinates are projected to an actual two-dimension ground plane so that a two-dimension ground image can be formed; data points of the same target are combined according to the position coordinates and distribution density of data points in the two-dimension ground image; the distribution center of the combined data points is taken as the positioned location of the target corresponding to the combined data points, and the target is tracked according to the positioned location. By the adoption of the target positioning and tracking method and device, the situation that one target is divided into multiple targets when the target is divided in the three-dimensional space can be avoided, computational burden is greatly reduced, and the target is accurately positioned and tracked.

Description

technical field [0001] The invention relates to the field of intelligent monitoring, in particular to target positioning and tracking technology. Background technique [0002] Positioning and tracking of human targets in complex scenes is a research hotspot in the field of computer vision, and it also has a large application space in the field of intelligent monitoring. At present, the most researched methods for positioning and tracking human targets include methods based on background modeling, and methods based on a large number of sample machine learning methods based on human body parts or entire regions. Since the above methods all use a single camera to capture scene images and analyze them The accuracy of positioning and tracking of human targets is greatly reduced due to the influence of light changes in the scene, the complexity of human targets, and shadows. Obtaining the depth image of the scene through the depth acquisition device, and using the depth image to ...

AI Technical Summary

Problems solved by technology

At present, the technology of using the depth image to locate and track the target mainly uses the calculation of projecting the foreground data of the target into the three-dimensional space, and then segmenting the target in the three-dimensional space. situation, low precision and large amount of computation

Images