Global motion estimation method based on space-time gradient extent and layering structure

Abstract

A global motion estimation method for time space gradient domain analyzing and layered construction based on light stream equation comprising, performing gauss lowpass and lower sampling processing to the two inputted image frames, then performing STGS preanalysis to get the important global motion information, applying these information into the parameter estimation module, seeking the optimum evaluation parameter using minimum iterating method, finally applying these results into the subsequent video compression, motion segment and video information categorization processing.

Description

technical field [0001] The present invention relates to a kind of method that carries out global motion estimation under two kinds of motion models, specifically, is a kind of analysis based on Special Temporal GradientScale (space-time gradient range of optical flow equation) applicable to video coding standard MPEG-4 / 7 and hierarchical global motion estimation methods. Background technique [0002] In general, the global motion estimation under any motion model is to use the estimated error ε(x, y) to minimize, so as to obtain the estimated motion parameters. If we simply sum the squares of the errors, and then calculate the parameters in the sense of the least square, the final estimation effect is too poor, and we often fall into the trap of local minimum, which leads to the failure of estimation. Therefore, a weighting function ω(ε) is generally generated to correct or improve parameter estimation. The design of this function reflects different estimation ideas, the mo...

AI Technical Summary

Problems solved by technology

Since the translation model cannot make a correct description of the rotation of the lens or the adjustment of the focal length, the affine model has a large error in the estimation of large translational motions, and the higher-order model has too much computational complexity.

Images