RESEARCH

Image & Vision

Investigates Image data fusion techniques that combine image and track data from multiple sensors to achieve improved accuracies and more specific inferences than could be achieved by the use of a single sensor alone. Our aim is to explore the state-of-the-art image processing algorithms for achieving effective data fusion as in:

Motion Mask based Robust Hand Tracking

Admin    2021-04-08    👁️ 207

1. Introduction
- Although hand tracking algorithm has been widely used in virtual reality hand HCI system, it is still a challenging problem
        in vision-based research area. due to the robustness and real-time requirements.

2. Main algorithm and principle
- motion mask based mean-shift tracking algorithm

1.    Building color probability map using color histogram of target.

2.    Motion detection by background subtraction.

3.    Color probability map updated by motion mask.

4.    Running mean-shift iteration on updated color probability map

5.    Prediction of search window using grey model

3. Experiment Results

- motion mask based color probability map

   - Hand tracking result