基于云平台CPU与GPU协同处理的光学卫星遥感影像正射融合方法
doi: 10.11728/cjss2025.05.2023-0069 cstr: 32142.14.cjss.2023-0069
-
于潇 女, 1980年10月出生, 现为北京通信与跟踪研究所副研究员, 主要研究方向为GPU在云平台环境中的应用. E-mail: yuxiao801026@126.com -
张一 女, 1997年4月出生, 现为北京跟踪与通信技术研究所研究实习员, 研究方向为人工智能和云技术在数据处理中的应用. E-mail: MinnieZ423@163.com -
张强 男, 1979年3月出生, 现为北京通信与跟踪研究所研究员, 主要研究方向为人工智能和云技术在数据处理中的应用. E-mail: 13581688087@126.com
作者简介:
通讯作者:
Optical Satellite Remote Sensing Image Orthographic Fusion Method Based on Coprocessing of CPU and GPU in Domestic Cloud Platform
-
摘要: 系统探讨了基于国产云平台调度下自主可控CPU和GPU协同处理的光学卫星遥感影像正射融合方法执行效率问题, 通过数据流配置、中间数据存储访问优化等手段进一步提高了该方法执行效率. 在云平台调度下, 使用飞腾S2500和英伟达A100对高分二号卫星多光谱影像进行正射融合的试验, 结果表明, 该方法可很大程度提高光学卫星遥感影像正射融合效率, 与传统X86架构CPU与GPU协同的正射融合算法相比, 加速比为14.3倍以上, 数据处理时间压缩至8.4 s内, 其中GPU运算耗时仅1 s, 可满足并优化大数据量的光学卫星遥感影像快速正射融合的要求.
-
关键词:
- 正射融合 /
- 国产云平台 /
- CPU和GPU协同处理 /
- 数据流配置 /
- 存储访问优化
Abstract: The processing efficiency of optical satellite remote sensing image orthographic fusion method based on coprocessing of CPU and GPU in domestic cloud platform is discussed systematically and is improved by data flow configuration and the intermediate data storage access optimization. The Phytium S2500 and NVIDIA A100 are used in the cloud platform to do the orthographic fusion experiment. The experiment results show that the method can greatly improve the fusion efficiency of optical satellite remote sensing image, and the acceleration ratio is more than 14.3 times of the traditional X86 architecture CPU and GPU collaborative orthographic fusion algorithm., and the corresponding processing time is reduced to less than 8.4 s, and the GPU operation time is only 1 s, which can meet the requirements of rapid orthographic correction of the large data of optical satellite remote sensing image. -
图 2 数据流配置和存储策略优化流程
Figure 2. Flowchart of data flow configuration and storage policy optimization
图 3 中国江西省某地多光谱与全色谱段数据对比
Figure 3. Comparison of multispectral with whole-chromatographic data from a place in Jiangxi Province
图 4 ENVI软件单独进行正射校正的输出图像对比
Figure 4. Comparison of output images for orthographic correction by ENVI software alone
图 6 ENVI软件单独进行局部正射校正的输出图像对比
Figure 6. Comparison of the output image of ENVI software for local orthographic correction alone
表 1 多组对照实验结果
Table 1. Results of multi-group controlled experiments
实验组 服务器配置及处理方式 是否MPI优化 正射校正与Gram-Schmidt
融合平均总耗时/s1 X86架构服务器CPU 是 197 2 国产云平台调度下的英伟达A100和飞腾S2500国产ARM架构
服务器(CPU和GPU协同处理)否 14.5 3 国产云平台调度下的英伟达A100和飞腾S2500国产ARM架构
服务器(CPU和GPU协同处理)是 8.4 
下载: 导出CSV
表 2 通用图像质量评价
Table 2. General image quality evaluation
谱段 通用图像质量评价指标 蓝谱段 0.884 341 绿谱段 0.881 685 红谱段 0.900 098 近红外谱段 0.904 558
下载: 导出CSV
表 3 结构相似性
Table 3. Structural similarity
谱段 结构相似性 蓝谱段 0.905 893 绿谱段 0.898 183 红谱段 0.907 704 近红外谱段 0.910 948
下载: 导出CSV
表 4 亮度均值
Table 4. Average brightness
谱段 MSS均值 融合结果均值 蓝谱段 321.998 567 321.770 95 绿谱段 258.338 339 258.186 605 红谱段 173.628 840 173.585 835 近红外谱段 389.297 872 389.379 371
下载: 导出CSV
表 5 光谱相关系数
Table 5. Spectral correlation coefficients
谱段 光谱相关系数 蓝谱段 0.884 392 绿谱段 0.881 768 红谱段 0.900 149 近红外谱段 0.904 606
下载: 导出CSV
-
[1] 方留杨, 王密, 李德仁. CPU和GPU协同处理的光学卫星遥感影像正射校正方法[J]. 测绘学报, 2013, 42(5): 668-675FANG Liuyang, WANG Mi, LI Deren. A CPU-GPU co-processing orthographic rectification approach for optical satellite imagery[J]. Acta Geodaetica et Cartographica Sinica, 2013, 42(5): 668-675 [2] 方旭东. 面向大规模科学计算的CPU-GPU异构并行技术研究[D]. 长沙: 国防科学技术大学, 2009FANG Xudong. Research on CPU-GPU Heterogeneous Parallel Technology for Large-Scale Scientific Computing[D]. Changsha: National University of Defense Technology, 2009 [3] 陈腾, 李微. 基于GPU和CPU协同加速的JPEG2000解码优化算法设计[J]. 电子世界, 2020(10): 99-101CHEN Teng, LI Wei. Design of JPEG2000 decoding optimization algorithm based on GPU and CPU co-acceleration[J]. Electronics World, 2020(10): 99-101 [4] 高原, 顾文杰, 丁雨恒, 等. 异构集群中CPU与GPU协同调度算法的设计与实现[J]. 计算机工程与设计, 2020, 41(2): 592-601GAO Yuan, GU Wenjie, DING Yuheng, et al. Design and implementation of CPU and GPU cooperative scheduling algorithm with heterogeneous clusters[J]. Computer Engineering and Design, 2020, 41(2): 592-601 [5] 张龙翔, 曹云鹏, 王海峰. 面向大数据复杂应用的GPU协同计算模型[J]. 计算机应用研究, 2020, 37(7): 2049-2053ZHANG Longxiang, CAO Yunpeng, WANG Haifeng. GPU collaborative computing model for complex applications in large-scale data processing[J]. Application Research of Computers, 2020, 37(7): 2049-2053 [6] 蒋艳凰, 杨学军, 易会战. 卫星遥感图像并行几何校正算法研究[J]. 计算机学报, 2004, 27(7): 944-951 doi: 10.3321/j.issn:0254-4164.2004.07.011JIANG Yanhuang, YANG Xuejun, YI Huizhan. Parallel algorithm of geometrical correction for satellite images[J]. Chinese Journal of Computers, 2004, 27(7): 944-951 doi: 10.3321/j.issn:0254-4164.2004.07.011 [7] 肖汉, 张祖勋. 基于GPGPU的并行影像匹配算法[J]. 测绘学报, 2010, 39(1): 46-51XIAO Han, ZHANG Zuxun. Parallel image matching algorithm based on GPGPU[J]. Acta Geodaetica et Cartographica Sinica, 2010, 39(1): 46-51 [8] 陈超, 陈彬, 孟剑萍. 基于GPU大规模遥感图像的几何校正[J]. 指挥信息系统与技术, 2012, 3(1): 76-80 doi: 10.3969/j.issn.1674-909X.2012.01.018CHEN Chao, CHEN Bin, MENG Jianping. Geometric correction of remote sensing images based on Graphic Processing Unit[J]. Command Information System and Technology, 2012, 3(1): 76-80 doi: 10.3969/j.issn.1674-909X.2012.01.018 [9] 侯毅, 沈彦男, 王睿索, 等. 基于GPU的数字影像的正射纠正技术的研究[J]. 现代测绘, 2009, 32(3): 10-11 doi: 10.3969/j.issn.1672-4097.2009.03.003HOU Yi, SHEN Yannan, WANG Ruisuo, et al. The discussion of GPU-based digital differential rectification[J]. Modern Surveying and Mapping, 2009, 32(3): 10-11 doi: 10.3969/j.issn.1672-4097.2009.03.003 [10] 杜慧. 吉林一号卫星影像融合及质量评价[J]. 测绘与空间地理信息, 2019, 42(5): 123-126 doi: 10.3969/j.issn.1672-5867.2019.05.038DU Hui. Image fusion and quality evaluation of Jilin-1 satellite[J]. Geomatics :Times New Roman;">& Spatial Information Technology, 2019, 42(5): 123-126 doi: 10.3969/j.issn.1672-5867.2019.05.038 [11] 肖昶, 余晓敏, 韩逸飞. 高分二号卫星影像融合技术研究[J]. 地理空间信息, 2018, 16(6): 13-16 doi: 10.3969/j.issn.1672-4623.2018.06.004XIAO Chang, YU Xiaomin, HAN Yifei. Research on image fusion technology of GF-2 satellite[J]. Geospatial Information, 2018, 16(6): 13-16 doi: 10.3969/j.issn.1672-4623.2018.06.004 [12] 吴敌, 汪红强, 邹同元. 基于GPU的遥感图像几何校正算法设计与实现[J]. 信息与电脑, 2020, 32(3): 38-40,43WU Di, WANG Hongqiang, ZOU Tongyuan. Design and implementation of geometric correction algorithm for remote sensing image based on GPU[J]. China Computer :Times New Roman;">& Communication, 2020, 32(3): 38-40,43 [13] 肖聆元, 李欣, 杨博. 基于CUDA架构的GF4影像快速正射纠正[J]. 测绘地理信息, 2019, 44(1): 74-78XIAO Lingyuan, LI Xin, YANG Bo. Quick orthographic rectification based on CUDA[J]. Journal of Geomatics, 2019, 44(1): 74-78 [14] 殷学永, 叶雨爽. 遥感影像融合方法的比较与评价[J]. 科技创新与生产力, 2020(6): 52-56,60 doi: 10.3969/j.issn.1674-9146.2020.06.052YIN Xueyong, YE Yushuang. Comparison and evaluation of remote sensing image fusion methods[J]. Sci-Tech Innovation :Times New Roman;">& Productivity, 2020(6): 52-56,60 doi: 10.3969/j.issn.1674-9146.2020.06.052 [15] 王福平, 陈超然, 吴翁慧. 基于GF-1号与ZY-3号遥感影像同源异源融合研究[J]. 江西测绘, 2021(2): 26-29WANG Fuping, CHEN Chaoran, WU Wenghui. Remote sensing image fusion of the same group and different groups based on GF-1 and ZY-3[J]. Jiangxi Cehui, 2021(2): 26-29 [16] 党源源, 王昕. CPU-GPU异构系统在光学遥感影像处理中的应用[J]. 红外与激光工程, 2020, 49(S1): 20200092DANG Yuanyuan, WANG Xin. Application of CPU-GPU heterogeneous system in optical remote sensing image processing[J]. Infrared and Laser Engineering, 2020, 49(S1): 20200092 [17] 胡怡, 陈道琨, 杨超, 等. 国产SW26010-Pro处理器上3级BLAS函数众核并行优化[J]. 软件学报, 2024, 35(3): 1569-1584HU Yi, CHEN Daokun, YANG Chao, et al. Many-core parallel optimization of level-3 BLAS function on domestic SW26010-Pro processor[J]. Journal of Software, 2024, 35(3): 1569-1584 [18] 齐健. NVIDIA更新Ampere架构, 全面提升GPU应用性能[J]. 智能制造, 2020(12): 30-31QI Jian. NVIDIA updates Ampere architecture to comprehensively improve GPU application performance[J]. Intelligent Manufacturing, 2020(12): 30-31 [19] 全吉成, 王平, 王宏伟. 计算机图形处理器加速的光学航空影像正射校正[J]. 光学精密工程, 2016, 24(11): 2863-2871 doi: 10.3788/OPE.20162411.2863QUAN Jicheng, WANG Ping, WANG Hongwei. Orthorectification of optical aerial images by GPU acceleration[J]. Optics and Precision Engineering, 2016, 24(11): 2863-2871 doi: 10.3788/OPE.20162411.2863 [20] 龚循强, 侯昭阳, 吕开云, 等. 结合改进Laplacian能量和参数自适应双通道ULPCNN的遥感影像融合方法[J]. 测绘学报, 2023, 52(11): 1892-1905 doi: 10.11947/j.AGCS.2023.20220541GONG Xunqiang, HOU Zhaoyang, LÜ Kaiyun, et al. Remote sensing image fusion method combining improved Laplacian energy and parameter adaptive dual-channel unit-linking pulse coupled neural network[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(11): 1892-1905 doi: 10.11947/j.AGCS.2023.20220541 [21] 许伟, 杨德芳, 陈李昊, 等. 多源遥感影像融合处理提取格拉丹东雪山区域特征信息[J]. 地质科技通报, 2024, 43(2): 370-385XU Wei, YANG Defang, CHEN Lihao, et al. Fusion processing of multisource remote sensing images for extracting characteristic information from the Geladandong Snow Mountain area[J]. Bulletin of Geological Science and Technology, 2024, 43(2): 370-385 [22] 田卫明, 刘富强, 谢鑫, 等. 基于GPU粗细粒度和混合精度的SAR后向投影算法的并行加速研究[J]. 信号处理, 2023, 39(12): 2213-2224TIAN Weiming, LIU Fuqiang, XIE Xin, et al. Research on parallel acceleration processing technology of SAR back projection algorithm based on two granularities and mixing precision of GPU[J]. Journal of Signal Processing, 2023, 39(12): 2213-2224 [23] 陈虎, 周鹏灵. 面向国产高性能众核处理器的编程模型[J]. 计算机应用, 2023, 43(11): 3517-3526CHEN Hu, ZHOU Pengling. Programming model for domestic high-performance many-core processor[J]. Journal of Computer Applications, 2023, 43(11): 3517-3526) [24] 杨靖宇. 摄影测量数据GPU并行处理若干关键技术研究[D]. 郑州: 信息工程大学, 2011YANG Jingyu. Research on GPU Parallel Processing of Photogrammetric Data[D]. Zhengzhou: Information Engineering University, 2011 [25] 李德仁, 王树根, 周月琴. 摄影测量与遥感概论[M]. 第二版. 北京: 测绘出版社, 2008LI Deren, WANG Shugen, ZHOU Yueqin. Introduction to Photogrammetry and Remote Sensing[M]. 2nd ed. Beijing: Surveying and Mapping Press, 2008 [26] 李俊杰, 傅俏燕. DEM对高景卫星影像正射校正的影响分析[J]. 地理空间信息, 2021, 19(9): 50-52,60 doi: 10.3969/j.issn.1672-4623.2021.09.012LI Junjie, FU Qiaoyan. Influence analysis of DEM on the orthorectification of SuperView-1 satellite images[J]. Geospatial Information, 2021, 19(9): 50-52,60 doi: 10.3969/j.issn.1672-4623.2021.09.012 [27] 梁文海, 陈琦, 杨承伶. 基于有理函数模型的GF-1影像正射校正分析[J]. 江苏科技信息, 2021, 38(7): 43-46 doi: 10.3969/j.issn.1004-7530.2021.07.014LIANG Wenhai, CHEN Qi, YANG Chengling. Orthorectification analysis of GF-1 image based on rational function model[J]. Jiangsu Science and Technology Information, 2021, 38(7): 43-46 doi: 10.3969/j.issn.1004-7530.2021.07.014 [28] 于海洋, 闫柏琨, 甘甫平, 等. 基于Gram Schmidt变换的高光谱遥感图像改进融合方法[J]. 地理与地理信息科学, 2007, 23(5): 39-42 doi: 10.3969/j.issn.1672-0504.2007.05.009YU Haiyang, YAN Baikun, GAN Fuping, et al. Hyperspectral image fusion by an enhanced Gram Schmidt spectral transformation[J]. Geography and Geo-Information Science, 2007, 23(5): 39-42 doi: 10.3969/j.issn.1672-0504.2007.05.009 [29] 童莹萍, 全英汇, 冯伟, 等. 基于空谱信息协同与Gram-Schmidt变换的多源遥感图像融合方法[J]. 系统工程与电子技术, 2022, 44(7): 2074-2083 doi: 10.12305/j.issn.1001-506X.2022.07.02TONG Yingping, QUAN Yinghui, FENG Wei, et al. Multi-source remote sensing image fusion method based on spatial-spectrum information collaboration and Gram-Schmidt transform[J]. Systems Engineering and Electronics, 2022, 44(7): 2074-2083 doi: 10.12305/j.issn.1001-506X.2022.07.02 [30] LI X H, LIN Y C, MENG M M, et al. Gram-Schmidt based hybrid beamforming for mmWave MIMO systems[J]. The Journal of China Universities of Posts and Telecommunications, 2016, 23(6): 53-59 doi: 10.1016/S1005-8885(16)60070-5 [31] LING F Y, MANOLAKIS D, PROAKIS J. A recursive modified Gram-Schmidt algorithm for least- squares estimation[J]. IEEE Transactions on Acoustics, Speech, :Times New Roman;">& Signal Processing, 1986, 34(4): 829-836 [32] SONG S J, SUN J, XU J Q. ARCA: A tool for area calculation based on GPS data[C]//26th International Conference on Database Systems for Advanced Applications. Taipei, China: Springer, 2021: 612-616 [33] 马冯, 孙旭, 高连如, 等. “高分四号”卫星正射校正精度分析[J]. 航天返回与遥感, 2019, 40(1): 74-82 doi: 10.3969/j.issn.1009-8518.2019.01.009MA Feng, SUN Xu, GAO Lianru, et al. Research on orthorectification accuracy of GF-4 satellite image[J]. Spacecraft Recovery :Times New Roman;">& Remote Sensing, 2019, 40(1): 74-82 doi: 10.3969/j.issn.1009-8518.2019.01.009 -
-
下载:
图(7) / 表(5)
计量
- 文章访问数: 248
- HTML全文浏览量: 28
- PDF下载量: 3
-
被引次数:
0(来源:Crossref)
0(来源:其他)
