Autonomous Motion Tracking for Dynamic Objects Using a Temporal Quad-Tree Algorithm

Tracking can be defined as calculating the trajectory of a moving object through a video. Tracking a moving object is an example of an automated process that needs to be done in real-time. Various tools are utilized for detecting moving objects such as sonar, laser, and computer vision. In recent years, with the massive improvements in the processing power of modern computers, computer vision techniques became a widely used tool for its tremendous advantages. In the proposed work, an object tracking technique named the Temporal Quad-Tree technique TQT is implemented to track a moving object within a series of images (a video). The proposed technique aims to enhance the computational time to detect a moving object. This technique will be implemented on a surveillance system, where time is an important issue. The system should track moving objects in an environment autonomously, especially when the presence of objects may be occluded for a certain period of time.