From: Object tracking system using a VSW algorithm based on color and point features
 | Step 1. Assign an initial search window to ROI. |
---|---|
 | Step 2. Compute the color probability distribution of the search window and generate a new search window by the meanshift algorithm. (The new search window has the same size but the different location as the initial search window.) |
 | Step 3. Extract the feature points for the SIFT algorithm around the new search window. |
 | Step 4. Make a rectangle with four points of the edge among detected feature points and assign the new search window to the rectangle. |
 | Step 5. Stop object tracking when the number of the feature points is under 4 or the location of the search window is not moved. Otherwise, repeat from Steps 2 to 4. |