Lossless computational reduction of full search algorithm in motionestimation using appropriate matching unit from image localization

摘要

To reduce the amount of computation of the full search (FS)algorithm for fast motion estimation, we propose a new and fast matchingalgorithm without degradation of predicted images as in conventional FS.The computational reduction without any degradation in the predictedimage comes from fast removal of impossible motion vectors. We obtainfaster removal of inappropriate motion vectors using efficient matchingunits from the localization of complex area in image data. In thispaper, we show three properties in block matching of motion estimation.Experimentally, we reduce unnecessary computations by about 30% with ouralgorithm compared with the conventional fast matching scan algorithm