| 摘要 |
For transcoding, a method for estimating a motion vector using a rate-distortion cost function and a transcoder using the method are disclosed. The transcoding method according to the present invention models a rate-distortion cost prediction function according to the dominant edge directivity of an input motion vector in an encoded input image and estimates an output motion vector having the minimum cost for rate-distortion by using the rate-distortion cost prediction function. [Reference numerals] (AA) Start; (BB) End; (S210) Decode an input image; (S220) Predict the main edge directivity of a rate-distortion cost for a motion vector in a macro block of the input image; (S230) Model a rate-distortion cost prediction function by using a weight variable matched with the main edge directivity; (S240) Estimate an output motion vector having the minimum cost from the rate-distortion cost prediction function |
