What is Motion Capture
Motion Capture is the process of recording the movement (position, rotation and acceleration) of objects, animals or people. It has a wide variety of applications in different fields, including animation and game development, life sciences, military, and virtual reality/simulation.
History of Motion Analysis
The first technology-based, systematic studies of motion were conducted by Eadweard Muybridge in the 19th century, before the advent of cinematography, and served for decades as the basis for understanding how humans and animals move.
By the early 20th century, when Walt Disney and others were pioneering animation, a detailed understanding of human motion – and the ability to reproduce it quickly and easily – became urgent. They adopted a technique called rotoscoping, invented earlier by Max Fleisher, whereby individual frames of movie film were traced onto animation cells as a means of speeding up the animation process.
By the time 3D computer animation came to dominate the entertainment industry, new techniques were sought to capture human motion in a more flexible way. But the need to capture accurate human motion was not limited to the field of animation. Doctors and scientists had pressing needs to better understand how patients moved in order to diagnose or study diseases and injuries. More sophisticated, flexible, computer-based motion capture technology was needed, and a number of technologies were developed to address these problems.
Modern motion capture systems have taken a variety of approaches to solving the problem of accurately tracking motion. Broadly, these fall into two general categories, optical and non-optical. Each has advantages and disadvantages in practice when applied to specific problem domains.
Optical Motion capture technology uses computer vision – the process by which digital images are processed by software to identify objects – in order to track motion. Typically, an array of video cameras is connected to a computer running specialized software. The object or person being tracked moves in sight of the cameras and specialized software analyzes each frame of video to try to create a 3D model of the action.
Marker-based Optical Motion Capture Systems
The field of computer vision is far from mature; replicating with technology what the human eye and brain can do so easily – quickly identify objects and track their motion – is a very difficult problem and still an active field of research. To make things simpler, most early approaches to optical motion capture made use of special markers attached to the subject. These markers were designed to be easily identifiable by image processing software. Typically, the markers are either highly reflective balls or else small bright lights that stand out from the background and are simpler for computer vision algorithms to identify.
While this allowed early motion capture systems to work, the downside is that it requires subjects to wear special suits with markers or to have the markers applied to their clothing or body. Precise application of the markers is essential, as the software is estimating the position of the subject (for example, the location and rotation of a person’s arm) by interpolating its position between a set of markers it is able to identify in the video image. If the marker slips, are misapplied, or become blocked by another object, the tracking will be inaccurate. Further, the time and overhead it takes to correctly apply markers makes many applications unfeasible. For example, tracking children within a clinical environment means spending a lot of time attaching markers to an active child before they can even step into the capture space. For animators or game developers, time spent in the setup and preparation stage takes away time from actually capturing motion, time not always available given the typically tight schedules of modern productions.
Markerless Motion Capture System
Markerless motion capture systems use advanced computer vision technology to identify and track subjects without the need for any special suits or markers. Without the aid of markers to provide hints to the image processing software, advanced algorithms are required, especially when the expectation is to track motion in real-time. This cutting edge technology is an active field of research at leading universities throughout the world. Organic Motion’s OpenStage 2 is the only commercially available professional markerless motion capture system on the market today.
There are many clear advantages to markerless motion capture. Because no special suits, markers or equipment are required, subjects can simply step into the capture volume to begin tracking. This means that animators and game developers can capture more accurate motion data in less time, and for a much lower total cost. It also makes many clinical and research applications much more practical. Because there is no special setup required, it is easier to track the motion of children or people whose condition makes the application of special suits difficult. The difficulty, however, is implementing accurate tracking algorithms that perform well enough for real-time use, without the aid of markers to provide hints to the software. This capability is the core of Organic Motion’s technology.
Non-Optical Motion Capture Systems
There have been a number of approaches to using non-optical systems to track motion. The most common of these is based on inertial and flex sensors. The idea is that a subject wears sensors that measure rotation, acceleration and flexion, and transmit the data to a computer that analyzes the results. For example an actor could wear inertial sensors on her limbs and flex sensors on her joints to capture data about how she moves. This has the advantage of not requiring complex computer vision technology to gather accurate data about relative movement. However, its limitation is that it only provides data relative to its last position, making the capture of multiple actors problematic. Also, it has the same problems as marker-based optical systems, namely that it requires extra setup and that the subject wear special equipment in order to be tracked.
Accurate motion capture is essential in several industries ranging from animation and game development to life sciences and medicine. While methodical studies of human and animal motion date back at least to the 19th century, the advent of modern computer-based motion capture system has recently provided the ability to capture accurate 3D motion in new and flexible ways. A variety of technologies have been used to meet this objective, including marker-based optical systems and systems based on inertial sensors. Newer technologies that provide markerless motion capture based on advanced computer vision algorithms are an open field of research at leading universities, and are available commercially today in Organic Motion’s OpenStage 2.
Schedule a Demo »
See how OpenStage 2 can change all you knew about motion capture in a 30 minute web seminar
Media and Product Information
Organic Motion OpenStage 2 Demo Video
"Organic Motion's technology improves every step of the animation process."