加权边沿自适应的场内插值去隔行方法*

为实现隔行扫描图像到逐行扫描图像的扫描格式转换，提出了一种基于空间加权和边沿自适应的场内插值去隔行算法。该算法包括LRV（left right vertical）评估和边沿自适应插值。相比传统的去隔行算法，LRV评估可以有效地消除伪边沿插值；在边沿自适应插值中增加预插值判断进行插值方向初步估计，可以抑制反向插值错误；在边沿方向搜索中增加空间权值，可以提高插值的准确率。实验结果表明，该算法在处理图片和视频流时的图像质量均要优于传统的算法。

De-interlacing with weighted edge adaptive intra-field interpolation

To implement the scanning format conversion from interlaced image to progressive image, this paper proposed a high performance de-interlacing algorithm based on spatial weighted and edge adaptive. The algorithm included two steps: LRV evaluation and edge adaptive interpolation. Compared with previous algorithms, LRV evaluation could remarkably reduce the fake edge interpolation, the prejudgment of edge before edge adaptive interpolation could suppress the reversal interpolation error, considered spatial weighting in weighted edge evaluation could reduce the edge interpolation error. Extensive simulations conducted for images and video sequences show the efficacy of the proposed algorithm over the previous algorithms in terms of the objective and subjective image quality.