Image restoration based on level lines
Ling Qi, Niu Feng, Liu Zhengkai · 2004
Images in nature can rarely be whole seen and they are occluded by something else. To get recovery of images, it need image restoring technology. Traditional methods are based on interpolation, which is not good if the occlusion is big. This paper is based on an existing theory and realizes a variational method: image restoration based on level lines. This approach simulates restoring ability of human vision, which can reconstruct edge features in images. Experiment results show this method can restore the information in big occluded area fairly well, moreover it spends time quite little. This method can be used to restore old photos and applied into image compression.