| 摘要 |
An image processing method includes: computing difference values between a boundary equivalent to an outer edge of a synthesis target region of an Nth frame of a video image, and a boundary equivalent to an outer edge of a synthesis target region of an N−1th frame of the video image; generating a difference image such that whilst pixel values at a boundary of the difference image maintain the difference values, a region inside the boundary of the difference image matches a specific gradient condition; and deriving, an image that is the generated difference image added to the synthesized result image derived for the N−1th frame that is equivalent to a result of synthesizing a synthesis image onto the synthesis target region in the N−1th frame, and substituting the synthesized result image derived for the Nth frame for an image of the synthesis target region in the Nth frame. |
