The use of object tracking in visual SLAM

This paper gives a research work in solving visual SLAM via a target-oriented tracking approach. The target-oriented tracking in this context refers to those object tracking approaches such as the well-known MeanShift but the approaches based on state-space or prediction are not included. The target-oriented tracking generally locks a dynamic or specific object which is not always available in a natural scene of mobile robots. In return, it gives us an inspiration to solve visual SLAM by tracking a region of interest (ROI) instead. By doing so, indirect methods relied on feature points can be applied to a local trajectory. Build on it, a semi-direct approach is finally proposed in this work which introduces a new way to solve visual SLAM. An experimental result is given to show how this idea works.