论文部分内容阅读
高分辨率傅里叶变换成像光谱仪具有高空间分辨率和高光谱分辨率的特点,但光谱重建时间冗长。通过对傅里叶变换光谱重建流程分析,为研制的1024pixel(光谱维)×1024piexl(像宽)×1024piexl(像高)高分辨率紫外傅里叶变换成像光谱仪的数据立方体反演,设计了一种并行优化算法。实验表明,在6核处理器上对512M和2G的数据立方体进行变换,时间分别只需88.33s和489.75s,加速比分别为3.70和3.04,大幅度提高了运算效率。如将该算法应用到更多内核处理器上,可得到更高的加速比和更少的运算时间。
High-resolution Fourier transform imaging spectrometers have the characteristics of high spatial resolution and hyperspectral resolution, but spectral reconstruction time is lengthy. Based on the Fourier transform spectral reconstruction flow analysis, a data cube inversion of the 1024 pixel (1024 × 1024 pixels × 1024 pixels) high resolution UV Fourier transform imaging spectrometer was developed. Parallel optimization algorithm. Experiments show that the data cube of 512M and 2G are transformed on the 6-core processor with the time being only 88.33s and 489.75s respectively, and the speedup ratios are 3.70 and 3.04 respectively, which greatly improves the operation efficiency. If the algorithm is applied to more core processors, get a higher speedup and less computing time.