This object tracking camera technology consists of an algorithm that is trained to track a given object in an offline manner by using information from the recorded video to optimize tracking. Current technologies to track objects either use sound localization or magnetic or light sources to track the object. This technology can be trained to recognize and track an object feature (e.g. a face). This video surveillance and analytics technology is useful for security applications such as investigating theft and sabotage for business and homeland security purposes.
Online tracking algorithm improves accuracy
This object tracking algorithm is first provided parameters of the object to be tracked. The algorithm then generates a position estimate of the object in the image to track the object. Tracking is optimized in an offline manner using current and future frames. This technology is in high demand in the surveillance market with a compound annual growth of 20% and revenue projection of $ 19 billion in 2017. Moreover, this technology may also be of demand in the personal photography market, recording sports events and defense applications.
- Surveillance in security
- Record lectures, talks etc.
- Customer tracking in retail stores
- Personal video recording (i.e. a ballet dancer practicing)
- Hollywood filming rehearsals
- Accurate tracking
- Less sensitive to occlusions, background noise and deformations