Motion video coding systems with motion vector detection

Abstract

A motion vector detecting circuit and a motion video compressing system are disclosed. The detecting circuit detect a vector which direct from a spatial location of a current macroblock to a spatial location of a reference macroblock. The detecting circuit includes a first extracting circuit, a second extracting circuit and a searching circuit. The first extracting circuit extracts upper bits from each pixel data of the current macroblock. The second extracting circuit extracts upper bits from each pixel data of the reference frame. The searching circuit searches the upper-bits-extracted reference frame for a macroblock having the strongest correlation with the current macroblock by comparing the upper-bits-extracted current macroblock with the upper-bits-extracted reference frame and outputs the spatial location of the reference macroblock. The motion video compressing system includes the motion vector detecting circuit, a transform-coding circuit, a subtracter, a comparator and a controller. The transform-coding circuit codes inputted data into compressed. The subtracter calculates a difference between each pixel data of the current macroblock and each corresponding pixel data of the reference macroblock directed by the motion vector. The comparator compares an amount of data being obtained by the transform-coding the current macroblock with an amount of data being obtained by the transform-coding the differential macroblock. The controller controls the transform-coding circuit such that the transform-coding circuit outputs which of the compressed current macroblock and the compressed differential macroblock has a smaller amount of data.