An area/time-efficient motion estimation micro core

An efficient microarchitecture for motion estimation is proposed. It achieves better time and area performance than the existing structures. Through pipelining and effective manipulation of two´s-complement arithmetic, the adder complexity is kept to its lowest, while speed for a combined subtraction, absolution, and accumulation operation is made as fast as a carry-save addition. Together with a new DCT (discrete cosine transform) algorithm, the microstructure is further expanded and tailored to facilitate efficient execution of other video operations, such as DCT and filtering operations