Vision-based real-time general movement capturing method and system for human body

Abstract

The invention discloses a vision-based real-time general movement capturing method and system for a human body. The vision-based real-time general movement capturing system comprises a multi-camera synchronous acquisition unit used for acquiring interested three-dimensional space scene video streaming, a real-time three-dimensional body and three-dimensional movement flow reconstructing unit used for reconstructing the three-dimensional body and whole-body three-dimensional movement flow information of the human body object in real time, a whole-body pose estimating and tracking unit used for estimating and tracking the three-dimensional general movement pose of the human body in real time, and a display unit using movement data to drive virtual characters in real time, wherein the display unit is used for using captured skeleton pose data to drive virtual characters in real time and displaying in a display device. The vision-based real-time general movement capturing method and system for the human body use a non-contact computer vision method, the real-time acquisition for the whole-body skeleton pose data and the real-time virtual character drive can be realized without an extra wear marking device, and compared with common action capturing technology and equipment, the vision-based real-time general movement capturing method and system for the human body are easy and convenient to realize.

Description

technical field [0001] The invention relates to the field of electronic information science, in particular to a vision-based real-time whole-body motion capture method and system thereof. Background technique [0002] Human motion analysis is one of the hot issues in computer graphics and computer vision research. The reason for this is that human motion data are used in many fields of entertainment and human-computer interaction. For example, in many video games or computer animations, human body motion data is often used to drive virtual characters, making the movements of virtual characters natural and lifelike; tracking and recognition of human body motions such as gestures can convert human body postures, gestures and other actions into certain computer commands, as a way of human-computer interaction; the third major application is in medical and sports, such as gait analysis is the observation and analysis of human body movements, body mechanics, and muscle activity ...

AI Technical Summary

Problems solved by technology

Although there are various multi-camera approaches, few of them can match the performance of traditional marker-based motion capture techniques, which are far worse in terms of accuracy, stability, and real-time performance of human tracking.

Images