how do I extract the transformation between frames when I using vision.pointtracker.

I have a video of falling object. I want to track the features (I used Harris Corners) in a ROI (this work ok). I am tracking using Lucas Kanade Tomasi algorithm in vision.pointtraker. How do I extract the tranformation between frames in the region I am tracking? . I want to later use this tranformation to stabilize the video base on the tracked ROI

[Pastor Villalpando]

Thanks a lot.

Answer 1

[Pratyush Roy]

Hi,

The estimateGeometricTransform function available in Computer Vision Toolbox helps to estimate the geometric transformation between two sets of matched points(corner points in this case). The documentation for the function can be obtained here:

https://www.mathworks.com/help/vision/ref/estimategeometrictransform.html

Hope this helps!