Normalized vegetation index parallel extraction method based on OpenACC

Abstract

The invention provides a normalized vegetation index parallel extraction method based on OpenACC. The method comprises the following steps: step 1, converting image data of a Mat matrix storage structure into image data of a sequential storage structure; 2, reading image data, and copying the image data to an acceleration device (GPU); 3, dividing all operations of the normalized vegetation indexextraction algorithm into a parallel operation and a non-parallel operation, performing parallel processing on the parallel operation based on OpenACC by adopting acceleration equipment, and processing the non-parallel operation by adopting a host, wherein the parallel operation comprises a histogram calculation operation, an image stretching operation, a histogram equalization operation, a normalized vegetation index calculation operation and an image binarization operation. According to the method, the characteristics of the vegetation topic indexes can be combined, and the normalized vegetation indexes can be rapidly and normatively extracted based on the OpenACC by utilizing the acceleration equipment in parallel.

Description

technical field [0001] The invention relates to the field of remote sensing application acceleration, in particular to an OpenACC-based parallel extraction method for normalized normalized vegetation index. Background technique [0002] The Normalized Difference Vegetation Index (NDVI) plays a vital role in agricultural remote sensing, and it is suitable for the dynamic monitoring of vegetation and the monitoring of temporal and spatial changes of urban vegetation coverage. At present, my country has established 30 high-resolution earth observation system data and application centers, and there is a huge amount of remote sensing satellite data that needs to be processed every year. Accelerating the processing of remote sensing big data has become a major issue in the process of remote sensing applications. [0003] Remote sensing image processing based on large-scale clusters is currently the mainstream processing mode. However, how to mine the data computing power of a sin...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that the traditional remote sensing image processing algorithm fails to propose a standardized parallel model for the acceleration process of vegetation index extraction when extracting normalized vegetation index, the present invention provides a parallel extraction of normalized normalized vegetation index based on OpenACC The method can combine the characteristics of the separation of serial and parallel operations in the vegetation index extraction process, and quickly and standardly extract the normalized normalized vegetation index in parallel.

Images