| 摘要 |
<p>A desired image is produced with less blur than an input image by correcting the input image, which contains multiple blurs, without using multiple images. An image correction device (100), which produces a desired image with less blur than an input image by correcting the input image, comprises an adaptive segmentation unit (110), which divides an input image into a plurality of adaptive segments according to the pixel values of the pixels that constitute the input image, a PSF interpolation unit (120), which uses a PSF that is calculated for each of the multiple divided segments and that shows the blur characteristics of an image, to interpolate the PSF for pixels located between representative pixels, which are pixels that represent each of the various multiple segments, and an image correction unit (130), which uses the interpolated PSF to correct the input image, thereby producing the desired image.</p> |
