Principle of optical motion capture system

Time:2022-5-7

Nokov measurement optical motion capture system is a motion capture system based on the principle of infrared optics. Compared with the positioning means such as inertial principle motion capture system and GPS positioning system, it has the characteristics of high precision, low delay, strong real-time performance and is mostly used in indoor scenes. The system establishment process can be divided into three parts: system construction, data acquisition and transmission, and data identification and processing.

10g.jpg

1. System construction

1.1 site construction

A set of optical motion capture system is composed of infrared optical lens, motion capture software, reflective mark points, Poe switch, cable, calibration frame, tripod and other lens fixing devices.

1g.jpg

Firstly, the infrared optical lens is arranged around the site through tripod, fixture and other lens fixing devices to ensure that the lens field of view can cover the capture area, and then all lenses are connected to Poe switch through network cable. The lens is powered and transmitted through Poe switch, and connected to the motion capture software in the computer. After the software starts, first connect the action capture lens in the real-time mode operation on the page.

2g.jpg

1.2 site calibration

 

After the software and hardware of the system are built and interconnected successfully, the next step is the calibration of the site, which is divided into L-type calibration and T-type calibration. Its function is to establish XYZ coordinate system for the motion capture area and calculate the position and attitude of each lens in the coordinate system. Only after the calibration is completed can the three-dimensional coordinate data of each marker point in the field be correctly obtained.

The L-shaped calibration is completed by placing the L-shaped calibration rod in the center of the site and setting it in the software. There are two purposes: first, determine the unified coordinate system. Through the positioning of the four points on the L-shaped rod, the system can distinguish its long axis and short axis, so as to define the orientation and origin position of the world setting axis. Secondly, this process can give an initial parameter to the lens that sees the L-shaped rod as the initial value of the subsequent parameter optimization.

The function of t-calibration is to provide enough data for each lens, so that it can carry out an iterative optimization of parameters on the basis of the original initial value. In this process, the software is in the T-shaped calibration mode, the operator swings the T-shaped rod in the field, and the lens captures a large amount of data in real time.

2 data acquisition and transmission

2.1 data acquisition

9.jpg

After the calibration is completed, the spatial data of the captured object can be obtained. A reflective mark point (a small silver gray ball coated with special reflective material) is pasted on the surface of the person or object to be positioned. The LED lamp on the action capture lens emits infrared light outward and receives the infrared light reflected from the reflective mark point at the same time. When multiple optical lenses “see” a mark point at the same time, the three-dimensional position of the mark point in space will be determined.

2.2 data transmission

5g.jpg

The position information of reflective mark points obtained by the lens needs to be transmitted to the computer in real time for data processing and use. In the optical motion capture system, all lenses are connected to the switch through the network cable. When the lens obtains the spatial position information of reflective mark points, these information will be transmitted to the switch through the network cable, and then uniformly transmitted to the connected computer by the switch, and received by the motion capture software in real time.

3 data identification and processing

3g.jpg

After the software obtains the three-dimensional spatial position of multiple reflective mark points, the next step is to recognize the object. Multiple reflective marker points are pasted on the surface of the same object, and the distance between these specific points is constant. Therefore, some points pasted on the same object are named, and the points are connected with lines to represent the relationship between the two points. This group of point names and connection information are operated and recorded as a group of markers in the software. When the object with this set of marketset information appears in the site, it is recognized as an independent object by the system.

4.jpg

Some human motion capture requires a large amount of point capture data, and there are special point models for selection. According to the provided point model, reflective mark points are pasted at the fixed position of the human body, and the points are identified, connected and bound with bones in the software.

When the system can recognize the captured object in real time, a complete optical motion capture system has been established. Next, the motion can be captured directly, and the captured model data can also be adjusted and corrected in the software in real time according to the effect. According to the needs of different fields, the optical motion capture system can also synchronously capture motion and force measurement data with force measuring platform and other equipment, and connect three-dimensional software to generate virtual characters.