Image compression and reconstruction method based on sparse representation

Abstract

The invention provides an image compression and reconstruction method based on sparse representation. Block division is carried out on the image by adopting a quadtree thought; a smooth area of a low-resolution image is divided into image blocks with the sizes of 64*64 and 32*32, a non-smooth image area containing detail information is divided into image blocks with the sizes of 4*4, and on the basis that the accuracy of dictionary training is guaranteed, the number of training samples is reduced, and the dictionary training time is shortened; when a decoding end reconstructs a decoded image,a smooth image block is directly reconstructed in a bicubic interpolation mode, an image feature block containing detail information is reconstructed in a sparse representation mode, and on the premise that the subjective quality of image reconstruction is guaranteed, the high-resolution image reconstruction time is shortened; image compression ratio is improved.

Description

technical field [0001] The invention relates to the field of image compression and reconstruction, in particular to an image compression and reconstruction method based on sparse representation. Background technique [0002] Image information has obvious advantages such as intuitive, vivid and concrete, and is one of the important ways for people to obtain information. However, high-quality images often occupy a large amount of storage space, and classic image compression technologies such as JPEG and JPEG2000 can compress the amount of data to the greatest extent while ensuring image quality as much as possible. However, with the emergence of high-definition and ultra-high-definition images, accompanied by the limitations of network bandwidth and storage media, traditional image compression technology sometimes cannot meet people's requirements for information acquisition. [0003] Due to the constraints of network bandwidth and computer storage space, low-resolution, comp...

AI Technical Summary

Problems solved by technology

However, high-quality images often occupy a large amount of storage space. Classical image compression technologies, such as JPEG and JPEG2000, can maximize the amount of data compression while ensuring image quality as much as possible.

However, with the emergence of high-definition and ultra-high-definition images, accompanied by the limitations of network bandwidth and storage media, traditional image compression technology sometimes cannot meet people's requirements for information acquisition.

Images