A cross-subset-guided measurement rate control method for residual block sets

Abstract

The invention relates to a method for controlling the measurement rate of a residual block set guided by a cross subset, comprising the following steps: dividing a target image into several blocks, classifying all the blocks, and classifying each block into a cross subset or a cross subset through the block set division. Residual block set; use the initial measurement rate to observe the intersection subset block by block in sequence, and count the observation value vector; set various optimized measurement rates for the remaining block set according to the statistics of the observation value vector in the cross subset, and use the optimal measurement Observation of the remaining block set block by block with high efficiency, and record the observation value; the observation value of each block is predicted in multiple directions block by block according to the spiral order, and the quantization and entropy coding of the residual are performed. The invention improves the overall quality of image observation while maintaining the average measurement rate.

Description

technical field [0001] The invention relates to the technical field of regulation and control of observation parameters in a compressed sensing communication system, in particular to a method for regulating and controlling the measurement rate of residual block sets guided by cross subsets. Background technique [0002] The compressed sensing communication system can break through the limitation of the Nyquist sampling law, so that the signal sampled at the sub-Nyquist rate can still be effectively restored, and the complexity can be transferred from the measurement end to the reconstruction end. The compressed sensing measurement terminal obtains low-dimensional observation data from the original high-dimensional signal, and the observation process is simultaneous signal acquisition and data compression. As a typical image compression sensing architecture, Quantized Block Compressive Sensing (QBCS) divides the target image into several blocks of the same size, each block co...

AI Technical Summary

Problems solved by technology

[0003] The inventors of the present invention have found that the current QBCS measurement terminal uses a single measurement rate to observe all blocks of the target image, ignoring the feature differences of different regions, so that image regions with little information are also assigned the same measurement rate, which is difficult Improving the observation quality of the target image

Images