Use of temporal redundancy of motion vectors for the increase of optical flow calculation speed as a contribution to real-time robot vision

Abstract

Robotic vision is an important aspect in robotic application and, especially, the measurement of real-time optical flow is a fundamental problem in the processing of image sequences and in robotic vision. We developed a new second order search method to improve the accuracy of optical flow and to still retain the advantage of the linear-time optical flow method by using temporal redundancy of the motion vectors of image sequences. The new method is a trade-off between accuracy and computational complexity of optical flow. This method will play an important role in real-time optical flow estimation using a personal computer for robotic vision.