VIDEO ENCODER CAPABLE OF DIFFERENTIALLY CODING SPEAKER'S VIDEO IMAGE AT PICTORIAL CALL AND VIDEO SIGNAL COMPRESSION METHOD USING THE SAME

摘要

PROBLEM TO BE SOLVED: To provide a video encoder capable of performing differential coding by dividing a speaker's image when making a pictorial call.

SOLUTION: The video encoder includes a motion estimation portion for estimating motions of an individual from a video signal and calculating a motion vector, a speaker region detection portion for detecting a speaker's region indicating a contour of the speaker from the motion vector, a discrete cosine transform portion for calculating a discrete cosine transform coefficient by performing discrete cosine transform on a video signal outputted from the motion estimation portion, a face region detection portion for generating a differential quantization table by detecting a face region of the speaker from the speaker region by using the discrete cosine transform coefficient and dividing the face region and a non-face region, an adaptive bit rate control portion for variably setting a quantization step size for performing quantization based on the speaker's region, and a quantization portion for performing quantization on the discrete cosine transform coefficient by the quantization step size and the differential quantization table.

COPYRIGHT: (C)2004,JPO&NCIPI