Non-subsampled contour wave transform optimization method based on parallel computing

Abstract

The invention discloses a non-subsampled contour wave transform optimization method based on parallel computing. The method comprises the following steps: (1) computing the number of the enabled GPUsand the number of the started CPU threads required for executing multi-scale decomposition and different-level direction decomposition in the NSCT algorithm according to the configuration situation ofthe GPU and the CPU and the actual computation amount allocated to each GPU; (2) performing parallelism analysis on the NSCT decomposition and reconstruction process, it is found that the image datacan be moved to the GPU and parallel processing of the convolution computation and computation result back storage process can be performed; and (3) using OpenMP and CUDA to perform NSCT decompositionand reconstruction in parallel. According to the method, the NSCT computation speed can be remarkably improved, the running time can be reduced and the practicability of the NSCT algorithm can be improved through the processes of parallel execution of data movement, pixel-level parallel computation convolution and the like.

Description

technical field [0001] The invention relates to a multi-scale geometric analysis method of an image, in particular to a non-subsampling contourlet transformation optimization method based on parallel operation, which belongs to the technical field of image processing. Background technique [0002] With the extensive research of digital image processing technology, multi-scale geometric analysis and other methods, a variety of transformation methods based on multi-scale geometric analysis have been developed one after another. The most representative transformation is the Contourlet transformation (contourlet transformation). The advantage of this transformation is that its basis functions are distributed in multiple scales and directions. It can effectively capture the edge contour of the image with a small number of transformation coefficients, making up for the Wavelet transform cannot efficiently represent the defects of image edge information. Therefore, Contourlet tran...

AI Technical Summary

Problems solved by technology

Therefore, the excellent performance of NSCT in the field of multi-scale geometric analysis of images has attracted the attention of researchers, but in practical applications, especially for high-resolution images, NSCT needs to perform multiple iterative filtering and two-dimensional convolution , causing the time overhead of the NSCT algorithm to be too large, it is difficult to use it efficiently in practical applications

Images