From: Object tracking system using a VSW algorithm based on color and point features
Given the distribution {q u }u = 1,...,mof a detected object and the location y0 of the detected object in the previous frame. | |
---|---|
Step 1. Compute the distribution p u in the current frame with y0 and compute the Bhattacharyya distance (= ρ) between p u and q u as follows:
| |
Step 2. Compute the weight (w i ) is defined as follow:
where b(x i ) is the color bin of the color at x i . | |
Step 3. Derive the new location y1 can be defined as follow:
where h is a window radius and g() is a kernel G. | |
Step 4. Compute the Bhattacharyya distance (= ρ) between p u and q u in the new location y1. | |
Step 5. While ρ|p(y0), q| <ρ|p(y1), q| Do y1 = 1/2(y0+y1). | |
Step 6. If ||y1-y0|| <ε Stop. Otherwise, y0 = y1 and go to Step 1. |