Distributed Image Compressive Sensing Reconstruction Method

Abstract

The invention provides a distributed compressive sensing reconstruction method for images. According to the method, the data volumes for image compression and transmission in a communication transmission system can be remarkably reduced, and the calculated amount and the data storage capacity of a decoding end are reduced to a greater degree. The distributed compressive sensing reconstruction method disclosed by the invention is realized by the following technical scheme: dividing an acquired original image into a plurality of sub images, and calculating a compression measurement value (the data compression encoding amount of a transmitting end for communication transmission) of each sub image by adopting compressive sensing projection; performing linear correlated weighing and fusing on compression measurement values (the data compression encoding amount of a receiving end for communication transmission) of the sub images to obtain a synthetic compression measurement value for image compressive sensing, and performing dimension extending and fusing processing on a sub measurement matrix according to a compression ratio setting value to obtain a comprehensive measurement matrix with relevance; finally, performing iterative data reconstruction on the synthetic compression measurement value by adopting an orthogonal matching pursuit reconstruction algorithm under the conditional constraint of a wavelet sparse matrix and the comprehensive measurement matrix to obtain a complete image without obvious overlapping edges and with high signal to noise ratio.

Description

technical field [0001] The invention relates to a high-efficiency data compression and decompression data reconstruction method for images under limited transmission bandwidth in the field of communication data link image transmission. Background technique [0002] With the equipment and use of high-resolution imaging reconnaissance cameras, there is a huge amount of images, which poses a challenge to the storage and transmission of image data on a limited platform. Therefore, effective sample storage, compressed transmission and high-quality decompressed image restoration of large amount of image data have become important technical problems. The existing image compression coding standards are all based on the traditional Shannon sampling theorem. Under the limitation of this theorem, the sampling rate of the vector must be greater than twice the bandwidth of the vector to accurately reconstruct the original vector. Due to the large amount of image data, more samples are r...

AI Technical Summary

Problems solved by technology

When compressing and reconstructing large pixel-level images, the overall image compression sensing method not only takes a long time at the encoding end, but also brings a huge amount of calculation and data storage at the decoding end, which is not conducive to the actual project on the decoding end equipment. Cost control; the block image compression sensing method successfully solves the problem of a large amount of calculation and storage for a single compression coding unit at the encoding end, but there is a large amount of data storage and calculation at the decoding end, and each sub-image compression measurement is ignored Strong correlation between the values, resulting in the loss of detail information in the stitched edges of the reconstructed image

Images