Detection method based on 3D skeleton modeling

Abstract

The invention relates to a detection method based on 3D skeleton modeling, which is characterized by comprising the steps of shooting a figure video by using a depth camera; carrying out face detection in the video; extracting a human body contour; carrying out verification on the human body contour, and judging whether the extracted contour is a human body contour or not, if yes, carrying out the next step, and if no, returning back to the step of face detection; carrying out image filtering and smoothing processing, and extracting a skeleton line by using a thinning algorithm; calculating the distance of parts of a human body, and establishing a probability distribution diagram; recognizing each part of the human body, and acquiring the coordinates of each joint point; connecting the joint points, and forming a complete 3D human body skeleton; outputting the coordinates of the 3D human body skeleton and hands; and tracking the human body contour, and providing information for the next frame. Compared with the prior art, the detection method provided by the invention has the beneficial effects that a person can be detected quickly, the human body skeleton line is acquired, the joint points of each person are positioned accurately, a 3D skeleton model is established, the calculation speed is high, the computation complexity is low, the detection method adapts to various complicated backgrounds, and each frame image only needs 5ms.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular, to a detection method based on 3D skeleton modeling. Background technique [0002] With the rise of depth cameras, air-operated computers based on depth cameras have increasingly entered people's lives. Robust human skeleton recognition has a wide range of applications in games, human-computer interaction, security, remote monitoring, and even medical treatment. With the introduction of depth cameras, new methods have been proposed in the field of human skeleton recognition. However, there is still no method that can completely display the outline of a person, generate a 3D skeleton model of the human body in real time, and obtain the 3D coordinates of each joint point. Contents of the invention [0003] The purpose of the present invention is to provide a detection method based on 3D skeleton modeling to solve the problem that the existing tec...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a detection method based on 3D skeleton modeling to solve the problem that the existing technology cannot completely display the outline of a person, and generate a 3D skeleton model of the human body in real time to obtain the 3D coordinates of each joint point

Images