3D Object Tracking
Select A Category
In today’s scenario, the technology for vision/visualization systems is reaching new heights. This has generated new ways to visualize and analyze images. Recent developments in 3D vision and display systems have created many opportunities to upgrade current practices in digital imaging and visualization.
On the analysis front, 3D is gaining a lot of momentum, especially 3D Object Tracking in real-time, which is finding many uses in Security & Surveillance, Medical, Automotive, Manufacturing, etc., where the following applications are in high demand:
- Tracking suspicious movements at public places through CCTV
- Tracking regions of interest and surgical tools during minimally-invasive surgery
- Tracking eyeballs, eyelids and eyebrows in Driver Drowsiness Systems
- Tracking industrial goods movement through robotic vision
The tracking software is a complex set of algorithms; different techniques are used to identify and track rigid or deformable objects. The detailed information about the position and orientation of the object in the scene is the key difference between 2D and 3D.
The commercial applications for 3D imaging technologies are projected to approach a market size of well over 1 billion USD by the end of 2011. So, seeing the importance of 3D object tracking in various industry applications, it can be assumed that it could cover 8-10% of the pie, and the growth is emphasized by the massive growth in stereoscopic cameras and 3D displays.
Some of the popular 3D object trackers are SynthEyes, 2d3 boujou, PFTrack and Maya (Unlimited) Live. Depending on usability, the prices range from 400 to 10,000 USD. In addition to these are Voodoo Camera Tracker, used for research, and OpenTL (Open Tracking Library), an open source tracking library.
Detailed information about an object in motion with respect to the background is highly useful information for a decision support system. With analysis of 3D data with preset parameters to generate alarm(s), along with suggested operator actions, a quick and accurate decision can be made.
3D imaging devices are gaining momentum in the market and upgrades are also happening at rapid pace, so it’s expected that next generation 3D imaging systems will include a tailored 3D object tracking feature. Moving forward, algorithms need to be developed to track several objects simultaneously, robustly and accurately in real time, i.e. Multiple 3D Object Tracking.