How the sampling time is choosen by calcOpticalFlowPyrLK??
- In the lucas kanade theory of optical flow measurement, instentanious velocity (vx,vy) of a point in an image is calculated at time t, given that the intensity will be same at time t+dt. But in calcOpticalFlowPyrLK, provide the new postion of the given input point at t+dt. So if they are using Euler method then how they select dt.
Or how they are calculating the position.
- What is the requirement to give input "nextImg" to the function calcOpticalFlowPyrLK() ??
Not sure but (Lucas–Kanade method) I think that Ix and Iy are the spatial derivatives across the first image and It is the derivative between the next and the current image.