Sk Telecom Company, Ltd.

Kyoung-Seok In

Chulwoo Kim

Min-Cheol Hong

Inseong Hwang

The video encoder in accordance with the present invention divides input image signal into macro blocks, estimates motion vectors of each macro block and finally encodes the input image signal. The encoder determines the number of macro blocks adjacent to a current macro block. If the number of adjacent macro blocks is equal to or more than two, the encoder calculates a motion vector of the adjacent macro blocks and selects a macro block that has the largest motion vector. Then it defines a least search area that the current adaptive motion vector can have, and compares the least search area with the motion vector of the largest adjacent macro block, and finally determines the largest value as the search area of the adaptive motion vector. Next, the encoder compares the search area of the adaptive motion vector with the search area of the user-defined motion adaptive vector to thereby determine the least value as the search area of the final adaptive motion vector.

Growing community of inventors

Seongnam-si, South Korea

Kyoung-Seok In