The analysis method of video camera's motion based on optical flow and slam

Abstract

This paper presents a method of video shot motion analysis based on LK optical flow method to solve the problem of film video shot motion of visual comfortable effect. A segmentation method based on footage of two class support vector machine classification is used. The optical flow method based on the hierarchical Lucas-Kanade in Pyramid combined with SLAM Algorithm to extract the motion information of the lens is used. Then the motion type of the shot is analyzed. Experimental results show that the method is not only able to achieve a good shot segmentation, but also the segmentation of the shot can be a good analysis of the results. Compared to the traditional optical flow method, the method has a higher accuracy. Finally it can analysis a movie, and get the distribution of different shots in the movie.