A novel accuracy assessment model for video stabilization approaches based on background motion

Abstract

In this paper, we propose a new accuracy measurement model for the video stabilization method based onbackground motion that can accurately measure the performance of the video stabilization algorithm. Undesired residualmotion present in the video can quantitatively be measured by the pixel by pixel background motion displacementbetween two consecutive background frames. First of all, foregrounds are removed from a stabilized video, and thenwe find the two-dimensional flow vectors for each pixel separately between two consecutive background frames. Afterthat, we calculate a Euclidean distance between these two flow vectors for each pixel one by one, which is regarded as adisplacement of each pixel. Then a total Euclidean distance of each frame is averaged to get a mean displacement foreach pixel, which is called mean displacement error, and finally we calculate the average mean displacement error. Ourexperimental results show the effectiveness of our proposed method.