摘要 |
According to an embodiment of the present invention, a weighted least squares-based intrafield de-interlacing algorithm comprises the following steps of: estimating a statistical model of an original image for a conditional probability model by Gaussian distribution; setting a local window, to which a weighted value of a preset magnitude is applied, in an interlaced image; and de-interlacing the local window based on an analysis of the statistical model. Therefore, the de-interlacing algorithm realizes high quality to estimate a fundamental original image and increases visible quality and a PSNR. |