What is the difference between Kinect motion capture and optical motion capture?

Kinect is a somatosensory camera with relatively low accuracy, which is only used in the peripheral devices of the Xbox 360 console, used for simple spatial positioning, and cannot be used in other fields that require motion capture, such as gait analysis, film and television animation, or education and scientific research. Optical motion capture technology is to complete the capture of the whole body action by tracking the target feature points from different angles by multiple high-speed cameras through the principle of computer vision, because of the high precision, it can be widely used in education and scientific research, medical and health, film and television entertainment and other fields.

This difference is like comparing the big brother of the past with the smartphone of today.

It is also used for "motion capture", and the boundary of Kinect is very narrow, and it was originally developed for game entertainment and C-end consumers. This means that it will not sell for a very high price and that the technical parameters are very limited (the so-called low precision).

Optical motion capture is "born extraordinary", and its initial development benefited from the needs of film and television shooting (many European and American blockbusters began to use motion capture technology early). With the development of the past twenty or thirty years, the current motion capture technology has been quite mature. Its application field is broader.

One of the major features of optical motion capture is the extremely high accuracy of capture. Behind this absolute advantage is a combination of hardware and software (expensive optical cameras and supporting software systems). This leads to its high overall cost, difficult operation, and the need for special personnel maintenance, so it is mainly oriented to the B-end or G-side.

This is also the main difference between it and Kinect.

Kinect is a somatosensory camera, and optical motion capture is based on the principle of computer vision, and multiple high-speed cameras track the target feature points from different angles to complete the capture of the whole body action. Optical motion capture devices are much more accurate than somatosensory cameras. The application field of Kinect is relatively narrow, and it can only be used in games at present, while optical motion capture systems can be used in drones, robot research, ergonomics, and film and television animation.

It can be used to capture the interface that drives the movement of the character in the development of the game engine, that is, it is used to play games, but it can't do industrial-grade capture in engineering, because the hardware camera of kinect has a low resolution, so the action accuracy is low and rough, so it can be reluctantly used after a long time and a large number of modifications to the action data, so the degree of coincidence with the original action is very large, and the captured action can hardly be used by regular engineering characters. The information of my profile QQ number is detailed in detail.

If you want to learn systematically, you can consider signing up for a live online class, and recommend CGWANG's online class. The teacher speaks carefully, you can watch it back after the class, and there are also the same type of recorded classes that you can learn for free (give away lifelong VIP).

This thing is generally a toy,First of all, the parameters of the kinect camera are not up to the standard of commercial capture hardware,You can only look at a rough idea,How to look like it,But once you really use it, you will find that the calculated data is unusable,It is impossible to adjust them all,Foreigners have also made a lot of such deceptive gadgets to sell,Body and expression tools are available,Personal play can be,But they can't reach industrial precision,You can only do an action expression approximately,Repair to meet the standard data will exhaust you, so it loses its meaning.

I'm a student of this major, but if you want to get 3D animation, this thing has to be calculated by seconds, and the animation industry uses seconds to calculate money, and some of them are tens, hundreds, thousands, foreign blockbusters, how many tens of thousands, one second, I personally think it's okay to make a short film and make it.

It can be commercially used. First of all, you need two computers, install win8 or above system and, because currently only one computer is supported to connect one, so two computers are needed. First of all, let's talk about the placement of the two Kinects, that is, the two Kinects are placed face to face, and the distance between them is about 4-6 meters.

Calibration is required before use. The method of calibration will be explained later. Then open IPI Recorder and select Kinect2 for Windows RecordVideo.

First of all, you need to save the background, and here you need to have no one in between. This is to be done on both Kinects. Then click this button on a computer.

This is done according to its laser positioning system. For example, the VR helmet of the Dapeng E3 positioning version adopts the Polaris laser positioning system, and the dual base stations realize 360° millimeter-level positioning around it, multi-directional laser cross scanning, large range, strong precision, and it is exciting to play VR games.