GRASS GSoC 2024 Parallelize Tools: Difference between revisions
m (Add category GSoC) |
Veroandreo (talk | contribs) (project won't be executed) |
||
| Line 13: | Line 13: | ||
;Abstract | ;Abstract | ||
This project aims to parallelize geospatial tools/algorithms to enhance our abilities to utilize and analyze large-scale geospatial and remote sensing data. Although many geographic information system (GIS) tools are widely used, they haven’t been parallelized yet. The serial computation, while functional for modest datasets, significantly restricts the processing time and scalability necessary for handling increasingly larger geospatial datasets. Leveraging the potential of modern computing hardware, we can parallelize tools to mitigate these limitations. Several tools could benefit from parallelization, such as r.texture, r.horizon, r.fill.stats, r/v.surf.idw, r.viewshed, v.to.rast, r.to.vect, r.grow.distance, etc. These tools can be speeded up by parallelization with OpenMP and MPI. Moreover, CUDA can be used for these tools to compute raster and image processing on GPU. | |||
;Note: | |||
Unfortunately, for administrative reasons out of the student's and GRASS GIS project control, this Google Summer of Code accepted proposal won't be executed. | |||
{{GSoC}} | {{GSoC}} | ||
Latest revision as of 20:05, 28 May 2024
Accepted Google Summer of Code 2024 project.
| Student Name: | Chung-Yuan Liang, Purdue University, West Lafayette, USA |
| Organization: | OSGeo - Open Source Geospatial Foundation |
| Mentor Name: | Huidae Cho, Vaclav Petras, Maris Nartiss |
| Title: | Parallelization of existing tools in GRASS GIS |
- Abstract
This project aims to parallelize geospatial tools/algorithms to enhance our abilities to utilize and analyze large-scale geospatial and remote sensing data. Although many geographic information system (GIS) tools are widely used, they haven’t been parallelized yet. The serial computation, while functional for modest datasets, significantly restricts the processing time and scalability necessary for handling increasingly larger geospatial datasets. Leveraging the potential of modern computing hardware, we can parallelize tools to mitigate these limitations. Several tools could benefit from parallelization, such as r.texture, r.horizon, r.fill.stats, r/v.surf.idw, r.viewshed, v.to.rast, r.to.vect, r.grow.distance, etc. These tools can be speeded up by parallelization with OpenMP and MPI. Moreover, CUDA can be used for these tools to compute raster and image processing on GPU.
- Note
Unfortunately, for administrative reasons out of the student's and GRASS GIS project control, this Google Summer of Code accepted proposal won't be executed.