Super-resolution reconstruction algorithm for JPEG compressed images

Study of super-resolution reconstruction algorithm for compressed images

Super-resolution reconstruction is a high-resolution image reconstructed from a low-resolution image of one or more images, such as a resolution 0.25m resolution image reconstructed from a remote sensing image of 4 1m resolution. There are very large applications in the military/civil.

At present, the super-resolution reconstruction method is divided into 3 categories: interpolation, learning-based, reconstruction-based approach. It has been studied much more now. However, most of the algorithms are used to study the normal image, and less for the compression image/video super-resolution reconstruction. Recent review of some of the literature. have been studied. Do some summary here.

Related literature:

1. Super-resolution from compressed video

2, Bayesian Resolution enhancement of compressed Video

3, Robustweb Image/video super-resolution

4, self-learning-based single Image super-resolution

of a highly compressed Image

Algorithm principle:

Document 1/2: Document 1/2 is written by the same person, the author published a number of articles on the issue, the method is basically similar. The idea is similar to the conventional super-resolution method based on reconstruction, which is to establish the observation model. The reconstruction cost function is established according to the data fidelity and regularization terms. The difference is that the data fidelity is somewhat different, due to the need to consider compressed image-specific quantization noise and rounding noise, rather than the usual if the Gaussian noise. In addition, the regularization items are different. It is necessary to consider the block effect of JPEG image, so in addition to the general constraints, the high frequency of block boundary is limited. See the literature for details.

Article 3: This document is relatively new, and it is 2010. The idea of the article is very well understood, first the JPEG compression image de-block effect, and then super-resolution. In the phase of block effect, the author uses adaptive Pde regularization method, and at the same time proposes an adaptive method to determine the regularization intensity. And in the reconstruction phase. is an example-based approach, and this approach is relatively mature.

In fact, such a two-step approach is unreasonable, it is not advocated for such treatment, but should be put together to consider, as in document 1/2.

Article 4: This document uses a sparse dictionary-related method. For the image to be rebuilt. First, the following sample is taken. and the original image and the lower sample image are used Bm3d processing to extract the low frequency, subtract from the original image and the lower frequency of the image (the author felt that the block effect is only in the high-frequency components, respectively, to obtain the original image and the lower-sample image of the high frequency is to train the dictionary, but also self-training dictionary).

After the high frequency is detected. The author proposes that high frequency is divided into block effect block/block effect block. Set up the dictionary d1/d2 respectively. And for a block effect dictionary D1. The author of the atomic Cluster, think that 1 parts of the atom is to contribute to the block effect, and some atoms do not contain block effect, then for the input block, I only use the atom without block effect of reconstruction, to achieve the purpose of the block effect.

In fact, the way to block effect. The author also wrote an article, efficient image/video deblocking VIA SPARSE representation, you can have a look. After looking at the better understanding of this article on the compression image super-resolution, is nothing more than the problem of dictionary construction.

Temporary on so many, not well written, you crossing will see, there are problems can be discussed again Oh!

！！！

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Research interests: Image compression, super-resolution, image denoising, image restoration and other image quality improvement techniques

Super-resolution reconstruction algorithm for JPEG compressed images