Wallis image enhancement method based on OpenCL (open computing language) architecture

Abstract

The invention discloses a Wallis image enhancement method based on an OpenCL (open computing language) architecture. The Wallis image enhancement method based on the OpenCL architecture is characterized in that during Wallis filtering parallel algorithm processing, a system adopts workgroup coarse-grain parallelism and worknode fine-grain parallelism; the steps of file reading, precomputation, data transmission, kernel start and result transmission are designed into a serial process; when scheduling is conducted on a multi-core CPU (central processing unit), image data, GPU (graphics processing unit) execution configuration parameters and image parameter information are transmitted to the equipment end, and multiplicative coefficients, additive coefficients and calculus of interpolation are designed into a parallel computing process suitable for a GPU architecture; when scheduling is conducted on a GPU, the multiplicative coefficients, the additive coefficients and new gray values of digital image pixel points are subjected to GPU parallel computing. The Wallis image enhancement method based on the OpenCL architecture has the advantages that the computation speed can be increased by two magnitude orders by a Wallis image enhancement parallel algorithm, so that computing speed can be increased remarkably.

Description

technical field [0001] The invention relates to a remote sensing mechanism and method, in particular to a Wallis image enhancement method based on OpenCL architecture. Background technique [0002] Aiming at the problem that Wallis image enhancement usually requires a large amount of calculation, it is difficult to perform real-time processing with traditional methods and leads to poor platform independence. In the past few decades, researchers at home and abroad have mainly focused on purchasing high-speed The single-core CPU is the computing core of servers, workstations and other equipment, and the application of multi-core CPU computing platforms adopts several methods such as the parallel processing structure of single instruction multiple data stream technology to speed up information processing. However, due to the constraints of manufacturing process technology and the limitations of the design goals of the CPU architecture itself, performance improvement soon encoun...

AI Technical Summary

Problems solved by technology

[0002] Aiming at the problem that Wallis image enhancement usually requires a large amount of calculation, it is difficult to perform real-time processing with traditional methods and leads to poor platform independence. In the past few decades, researchers at home and abroad have mainly focused on purchasing high-speed The single-core CPU is the computing core of servers, workstations and other equipment, and the application of multi-core CPU computing platforms adopts several methods such as parallel processing structure of single instruction multiple data stream technology to speed up information processing.

However, due to the constraints of manufacturing process technology and the limitations of the design goals of the CPU architecture itself, performance improvement soon encountered technical barriers. These traditional computing resources and computing acceleration methods are difficult to meet the needs of real-time processing of remote sensing images. Using the Wallis image enhancement algorithm The contradiction between the real-time processing of massive remote sensing data and the relatively lagging means of data processing and processing has become increasingly prominent and acute

The problems of incompatibility between hardware platforms and poor cross-platform portability of software are prominent

Images