An Important Frame Distinction Model of Stereoscopic Video Based on Content

Abstract

As different frame lost will lead to different distortion for stereoscopic video sequence, we take into account the temporal and spatial correlation of stereo video, error diffusion characteristics left and of right view frame, as well as recursion theory. Based on the above, we put forward a frame important distinction model of stereoscopic video based on content. The experimental results show that the model can be applied for stereoscopic video sequences with different motion intensity and disparity, accordingly, code ends can accurately estimate the importance of each frame for terminal perception. Finally, the applicability of the model is discussed.