Image enhancement based on matrix completion

In recent years, how to effectively complement preferable details to an image according to its local information is one of the research focuses in the field of image enhancement. For an image badly lack of local details, the key of image enhancement is to reconstruct the unknown original details in terms of the small amount of known information. To completely or approximately reconstruct an unknown signal by a small number of its known elements is a matrix completion problem in the sparse theory. This paper proposes an image enhancement algorithm based on matrix completion, which implements effective complements of local details to the local blurred image by solving the nuclear norm minimization problem with the method of singular value shrinkage iteration, and achieves image enhancement with fine subjective qualities for human vision.