A novel de-interlacing method based on locally-adaptive Nonlocal-means

This paper presents an efficient method for video de-interlacing based on Nonlocal-means (NL-means). In the proposed scheme, every interpolated pixel is set to a weighted average of its neighboring pixels. Weights of the pixels are calculated according to the radiometric distance between the surrounding areas of them and the pixel being interpolated. To calculate the weights, we need an estimate of the progressive frames. Therefore, we initially de-interlace frames using a simple and reliable edge-based de-interlacing method. We use steering kernel in NL-means to adapt it locally to the features of the image. Experimental results show the effectiveness of our method.