Apparatus and method for detecting motion vectors to half-pixel accuracy

Claim 1

A method for determining motion vectors between a current frame and a preceding frame of video signals to a half-pixel accuracy, wherein the current frame is divided into a multiplicity of search blocks of an identical size and the preceding frame is made to include a corresponding number of search regions, each of the search regions having M candidate blocks of said identical size, M being an integer greater than 1, which comprises the steps of: (a) calculating M first error functions, each of the M first error functions representing the difference between a search block in the current frame and each full-pixel resolution candidate block, included in the corresponding search region of the preceding frame; (b) selecting K full-pixel resolution motion vectors for the search block which yield smallest first error functions among said M first error functions, K being an integer greater than 1 and smaller than M; (c) obtaining M half-pixel resolution candidate blocks, in a half-pixel resolution search region corresponding to one of the K full-pixel resolution motion vectors; (d) calculating M second error functions, each of the M second error functions representing the difference between the search block in the current frame and each of said M half-pixel resolution candidate blocks; (e) repeating steps (c) to (d) for the remaining (K-1) full-pixel resolution motion vectors obtained in step (b), providing K×M second error functions; (f) selecting a half-pixel resolution motion vector for the search block, said half-pixel resolution motion vector yielding a second error function which is smallest among the K×M second error functions; and (g) repeating steps (a) to (f) until all the search blocks in the current frame are processed.

Claim 2

A method for determining motion vectors between a current frame and a preceding frame of video signals to a half-pixel accuracy, wherein the current frame is divided into a multiplicity of search blocks of an identical size and the preceding frame is made to include a corresponding number of search regions, each of the search regions having M candidate blocks of said identical size, M being an integer greater than 1, which comprises the steps of: (a) calculating M first error functions, each of the M first error functions representing the difference between a search block in the current frame and each full-pixel resolution candidate block, included in the corresponding search region of the preceding frame; (b) selecting K full-pixel resolution motion vectors for the search block which yield smallest first error functions among said M first error functions, K being an integer greater than 1 and smaller than M; (c) obtaining M half-pixel resolution candidate blocks, in a half-pixel resolution search region corresponding to one of the K full-pixel resolution motion vectors; (d) calculating M second error functions, each of the M second error functions representing the difference between the search block in the current frame and each of said M half-pixel resolution candidate blocks; (e) repeating steps (c) to (d) for the remaining (K-1) full-pixel resolution motion vectors obtained in step (b), providing K×M second error functions; (f) selecting a half-pixel resolution motion vector for the search block, said half-pixel resolution motion vector yielding a second error function which is smallest among the K×M second error functions; and (g) repeating steps (a) to (f) until all the search blocks in the current frame are processed.