What Is Visual Capture?

Motion capture is a tracker set at a key part of a moving object. Same meaning: Motion capture English Motion capture, referred to as Mocap. Technology involves data that can be directly understood by a computer in terms of dimensional measurement, positioning of objects in physical space, and orientation determination.

motion capture

2008 by

common

With the rapid development of computer software and hardware technology and the increase in animation production requirements, in developed countries, motion capture has entered the stage of practicality, and many manufacturers have introduced a variety of commercial motion capture equipment, such as MotionAnalysis, Polhemus, Sega Interactive, MAC, X-Ist, FilmBox, Nokov, etc. have been successfully used in many aspects such as virtual reality, gaming, ergonomics research, simulation training, biomechanical research, etc.

From a technical perspective, the essence of motion capture is to measure, track, and record the motion trajectory of an object in three-dimensional space. A typical motion capture device generally consists of the following parts:

· Sensor. The so-called sensor is a tracking device fixed at a specific part of a moving object. It will provide the motion capture system with position information on the movement of the moving object. Generally, the number of trackers will be determined with the detail of the capture.

· Signal capture equipment. Such devices differ depending on the type of Motion capture system, they are responsible for the capture of position signals. For mechanical systems, it is a circuit board that captures electrical signals. For optical Motion capture systems, it is a high-resolution infrared camera.

Data transmission equipment. Motion capture systems, especially Motion capture systems that require real-time effects, need to quickly and accurately transfer a large amount of motion data from the signal capture device to a computer system for processing, and the data transmission device is used to complete this work.

Data processing equipment. The data captured by the Motion capture system needs to be corrected, and a three-dimensional model must be combined after processing to complete the computer animation production. This requires us to use data processing software or hardware to complete this work. Both software and hardware, they all rely on the computer's high-speed computing power to complete the data processing, so that the three-dimensional model moves truly and naturally.

Technology one: mechanical motion capture

Mechanical motion capture relies on mechanical devices to track and measure motion trajectories. A typical system is composed of multiple joints and rigid links, and an angle sensor is installed in the rotatable joint to measure the change of the joint rotation angle. When the device moves, according to the angle change measured by the angle sensor and the length of the connecting rod, the position and movement trajectory of the end point of the rod in space can be obtained. In fact, the motion trajectory at any point on the device can be obtained, and the rigid link can also be replaced with a telescopic rod of variable length, and the change in length is measured with a displacement sensor.

An early type of mechanical motion capture device used joints and connecting rods with angle sensors to form a "adjustable attitude digital model". Its shape can simulate the human body, as well as other animals or objects. The user can adjust the pose of the model according to the needs of the plot, and then lock it. The angle sensor measures and records the rotation angle of the joint. Based on these angles and the mechanical size of the model, the model's pose can be calculated, and the pose data is transmitted to the animation software, so that the character model in it also makes the same pose. This is an early motion capture device, but it still has a certain market until now. A very vivid name has been given to this device abroad: "monkey".

An application form of mechanical motion capture is to connect a moving object to be captured with a mechanical structure, and the movement of the object drives the mechanical device, so that it is recorded by the sensor in real time.

The advantages of this method are low cost and high accuracy. It can measure in real time and allow multiple characters to perform at the same time. However, its shortcomings are also very obvious, mainly because it is very inconvenient to use, and the mechanical structure hinders and limits the performers' movements. The "monkey" is more difficult to use for real-time capture of continuous action. It requires the operator to constantly adjust the "monkey" pose according to the requirements of the plot. It is very troublesome and is mainly used for static shape capture and key frame determination.

Technique 2: Acoustic Motion Capture

The commonly used acoustic motion capture device consists of a transmitter, a receiver, and a processing unit. The transmitter is a fixed ultrasonic generator, and the receiver generally consists of three ultrasonic probes arranged in a triangle. By measuring the time or phase difference of the sound wave from the transmitter to the receiver, the system can calculate and determine the position and direction of the receiver.

This type of device has low cost, but has a large delay and lag in motion capture, poor real-time performance, and generally not very high accuracy. There must be no large occluded objects between the sound source and the receiver, and noise and multiple reflections. Noisy. Because the speed of sound waves in the air is related to air pressure, humidity, and temperature, corresponding compensation must be made in the algorithm.

Technology III: Electromagnetic Motion Capture

Electromagnetic motion capture systems are relatively common motion capture devices. Generally consists of a transmitting source, a receiving sensor and a data processing unit. The transmitting source generates an electromagnetic field that is distributed in a certain space-time law in space; the receiving sensors (usually 10 to 20) are placed at key positions on the performer's body, and move with the performer's movement in the electromagnetic field, and communicate with the data through cables or wireless The processing units are connected, as shown in Figures 2 and 3.

When a performer performs in an electromagnetic field, the receiving sensor transmits the received signals to the processing unit through a cable. Based on these signals, the spatial position and direction of each sensor can be calculated. Both Polhemus and Ascension are known for producing electromagnetic motion capture devices. The sampling rate of such systems is generally 15 to 120 times per second (depending on the number of models and sensors). In order to eliminate jitter and interference, the sampling rate is generally below 15Hz. For some high-speed sports, such as boxing, basketball games, etc., the sampling speed cannot meet the requirements. The advantage of electromagnetic motion capture is firstly that it records six-dimensional information, that is, it can obtain not only the spatial position but also the direction information, which is very valuable for some special applications. The second is fast speed and good real-time performance. When performers perform, the character model in the animation system can react at the same time, which is convenient for rehearsal, adjustment and modification. The calibration of the device is relatively simple, the technology is mature, the robustness is good, and the cost is relatively low.

Its disadvantage is that it has strict requirements on the environment, and there must be no metal objects near the performance venue, otherwise it will cause electromagnetic field distortion and affect accuracy. The allowable performance range of the system is smaller than that of the optical type, especially the cable has a greater limit on the performer's activities, and it is not suitable for more vigorous sports and performances.

Technology 4: Optical Motion Capture

Optical motion capture completes the task of motion capture by monitoring and tracking specific light points on the target. Common optical motion capture is mostly based on computer vision principles. In theory, for a point in space, as long as it can be seen by two cameras at the same time, according to the images and camera parameters taken by the two cameras at the same time, the position of the point in space at that time can be determined . When the camera continuously shoots at a sufficiently high rate, the motion trajectory of the point can be obtained from the image sequence.

A typical optical motion capture system usually uses 6 to 8 cameras arranged around the performance venue. The overlapping areas of these cameras 'fields of vision are the performers' range of motion. In order to facilitate processing, performers are usually required to wear single-color clothing, and put some special signs or light spots on key parts of the body, such as joints, hips, elbows, wrists, etc., called "

In the 1970s and 1980s, motion capture began to be used as a tool for analyzing photographic images in biomechanical research. As the technology matured, the technology began to expand to education, training, sports, computer animation, television, movies, video games And other fields. The user is provided with marker points at each joint, and recognizes movements by changing the position and angle between the marker points.

Motion capture systems include mechanical links, magnetic sensors, light sensors, acoustic sensors, and inertial sensors. Each technology has its own advantages, but no matter what technology, users will have some restrictions.

Optical uses optical perception to determine the real-time position and orientation of an object. Optical devices mainly include light-sensitive devices (receivers), light sources (transmitters), and controllers for signal processing. There are many kinds of light-sensitive equipment, such as ordinary video cameras and photodiodes. The light source can be ambient light or structured light. In order to prevent the interference of visible light, infrared light, laser light, etc. are usually used as the light source. Because light travels fast, the most significant advantages of optical devices are fast speed, high update rate, and low latency, which is more suitable for occasions with strong real-time performance and works well in a small range. The working principle of infrared passive optical motion capture: a capture space composed of multiple cameras. The near-infrared LED on the camera illuminates the reflective marking points on the target (the circle of light sources in the picture above is the near-infrared light source). Perform infrared imaging, extract the two-dimensional information of the marked points, and calculate the three-dimensional position information of the marker points through the spatial data fed back by the multiple cameras to the same marked point. The motion capture system will complete the continuous shooting of the performers' actions, image storage, Analyze, process and complete the real-time recording of the motion trajectory.

The position of the tracked object is obtained by blind inference by inertial method, that is, it is completely estimated by the internal movement system. The advantage is that there are no emission sources, no fear of blocking, no external interference, and unlimited working space. The disadvantage is the rapid accumulation of errors. ^[4]

The mechanical type is an older tracking method, which is composed of a link device. It is a relatively inexpensive system with high accuracy and short response time. It can measure the entire body movement of an object without delay, and it is not affected by external interference such as sound, light, electromagnetic waves. In addition, it can be combined with a force feedback device. The disadvantage is that it is bulky, inflexible, and inertia. Due to the limitation of mechanical connection, its working space is also restricted to a certain extent, and there is also a central zone in the working space that is inaccessible, commonly known as the dead end of the mechanical system, which prevents mechanical equipment from entering.

Electromagnetic type uses the strength of the magnetic field for position and orientation tracking. Generally includes transmitter, receiver, interface and computer. The advantage is that there is no occlusion problem, and other objects are allowed between the receiver and the transmitter, which allows the user to move around. Compared with other motion capture devices, it has a lower price, moderate accuracy, high sampling rate (up to 120 times / second), and a large working range (up to 60m), allowing multiple magnetic trackers to track the entire body movement, and increase To track the range of motion. The disadvantage is that it is susceptible to interference from electronic equipment and ferromagnetic materials, which may cause magnetic field deformation and errors. As the measurement distance increases, the error increases, the time delay increases (33ms), and there is small jitter.