Adaptive compressive sensing image obtaining method based on optimal truncation model

Abstract

The invention discloses an adaptive compressive sensing image obtaining method based on an optimal truncation model, and the method comprises the steps: carrying out the optimal truncation of a transform domain coefficient of an inputted scene image signal according to the difference of compressive sensing sampling rates and the optimal truncation model, and then carrying out the adaptive compressive sensing obtaining of the scene image signal after truncation, wherein an obtained image compressive sensing sampling value obtains a quantitative codon through the unique optimal quantification based on lookup mapping; and finally obtaining quantification output results. The method guarantees that the transform domain sparse degree of the inputted scene image signal is optimal through employing the optimal truncation model, thereby greatly improving the quality of a compressive sensing reconstruction image. The method achieves the high-efficiency processing of sampling data through the unique optimal quantification based on lookup mapping, and can save the energy and calculation resources of a system. The method can meet the requirements of application environments which exerts energy and computing complexity limits on an image obtaining system, such as a wireless multimedia sensing network, spatial image obtaining, mobile terminal image, and so on.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to an adaptive compression sensing image acquisition method based on an optimal truncation model. Background technique [0002] The image processing technology based on compressed sensing has changed the traditional image processing method - the system mode of complex encoding end and simple decoding end greatly reduces the consumption of energy and computing resources at the encoding end, and reduces the consumption of energy and computing resources. The burden is shifted to the relatively rich decoding end, which expands the application of image processing technology in multimedia sensor networks, spatial image acquisition, mobile terminal imaging and other fields. A traditional digital image processing system includes two cascaded modules: 1) an optical sensor array for collecting scene pixels, and 2) a data compression processor for reducing the data volume of the col...

AI Technical Summary

Problems solved by technology

However, this system structure cannot meet the needs of some application scenarios

First of all, in some application scenarios, digital image processing systems have energy supply constraints and limited computing resources; secondly, in some application scenarios, the cost and complexity of the system increase greatly with the increase of image spatial resolution

For this reason, digital imaging systems based on compressed sensing technology solve these problems by combining scene pixel sampling and data compression processes. However, in terms of compressed sensing image acquisition and subsequent data compression, as well as the energy consumption and computational complexity Requirements, how to use the existing system framework, combined with optimization technology, to realize the efficient acquisition and processing of digital images has become a prominent problem in the current system design

Images