Adaptive estimation of 3-D motion parameters and optical flow

The recovery of the 3-D motion parameters and structure of an object from its 2-D dynamic picture image is one of the most important problems in computer vision. First, we present a robust adaptive update law based on Riccati equation. Moreover, we design adaptive estimators of 3-D motion parameters and its optical flow from image sequences directly without differential operations. Finally, we demonstrate simulation results via Matlab/Simulink.