Computational Tool for Analysis of Strains Based on Optical Flow Approach

Abstract

At present, there is an increase in the quantity of computer vision method applications in the industry, robotics, in systems of virtual and augmented reality, geoinformation systems and so on. This leads to the need for the creation and development of efficient methods and algorithms for solving the problems of computer vision. In the paper, we present an approach to solve the problem of the finding of displacement vectors in the two consecutive video frames of the strain test process video. To analyse video we use the optical flow approach and propose improvements to feature extraction method of establishing of robust point correspondences between pair of key points in video frames.