دانلود رایگان مقاله ISI درباره طرح نقشه،داده های فضایی بزرگ، پردازش ابری و واحد پردازش گرافیکی
دانلود رایکان مقاله انگلیسی ISI با موضوع طرح نقشه برداری موازی داده فضایی بزرگ بر مبنای بردار
عنوان فارسی مقاله:
طرح نقشه برداری موازی داده فضایی بزرگ بر مبنای بردار: ادغام محاسبات ابری با واحد پردازش گرافیکی
عنوان انگلیسی مقاله:
Parallel map projection of vector-based big spatial data: Coupling cloud computing with graphics processing units
دانلود رایگان مقاله ISI با فرمت PDF:
مشاهده توضیحات کامل و خرید ترجمه فارسی با فرمت ورد تایپ شده:
بخشی از مقاله انگلیسی :
3. Methods
In this section, we present the design of the parallel computing framework for the accelerated map projection of vector-based big spatial data. We choose to use bare earth LiDAR data of North Carolina, USA as an example of vector-based big spatial data (see Fig. 1). This dataset covers the entire North Carolina, resulting in 13,596 sub-datasets (in file format). The LiDAR dataset is collected and released by the North Carolina Floodplain Mapping Program (see http://floodmaps.nc.gov). The original projection of the dataset is the Lambert Conic Conformal projection. About 230 GBs of storage space are needed to maintain the original LiDAR dataset. As an example, it takes about several hours of CPU-based sequential time (2.83 h for projection only; 11.46 h with the consideration of input/output operations; an advanced CPU is used) for re-projecting the entire dataset from Lambert Conic Conformal projection to Mollweide projection. Thus, a parallel framework that allows for efficient map projection is necessary and urgent if we are to transform these vector-based big spatial data. Thus, in this section, we first describe the entire architecture of the framework. Then, we discuss each component of the framework