Moving picture encoding apparatus and method

Claim 1

Apparatus for compressing a motion picture signal to provide a compressed signal, wherein the motion picture signal is divided into blocks of pixels, a moving vector is obtained with one-pixel precision between a matching block of a previous picture and a current block, a moving vector with half-pixel precision is derived from the moving vector with one-pixel precision, differences between pixel data of the current block and movement compensated picture data obtained in response to the movement vector with half-pixel precision are orthogonally transformed, and resulting transform coefficients are quantized with a predetermined quantizing step size to provide the compressed signal, the apparatus comprising: parallelizing means for converting serial pixel data having a first word length into parallel pixel data having a second word length greater than the first word length; interpolation data generating means for performing interpolation on the parallel pixel data to generate interpolation data with respect to each pixel in an area designated by the moving vector with one-pixel precision; matching block generating means for providing the interpolation data output from the interpolation data generating means as interpolation data in plural matching blocks displaced by half a pixel from a matching block designated by the moving vector with one-pixel precision; difference determining means for obtaining, for each of the plural matching blocks from the matching block generating means, a difference between the interpolation data of the matching block and the pixel data of the current block; and means for determining the moving vector with half-pixel precision, the moving vector with half-pixel precision corresponding to a displacement quantity of the one of the matching blocks determined to have the least difference.