Skip to content

Opensource GIS Tool leveraging NVIDIA CUDA and pyCuda for fast Raster Analysis

Notifications You must be signed in to change notification settings

aFuerst/PyCUDA-Raster

Repository files navigation

Lembo-REU-2016

Opensource GIS project leveraging NVIDIA CUDA and pyCuda

Students: Alex Fuerst, Charlie Kazer, Billy Hoffman

Synopsis

A command line tool and plugin for QGIS that leverages a CUDA enabled GPU for raster calculations. Currently supports calculating slope, aspect, and hillshade from an elevation file.

Status

Command line tool, and stand-alone GUI work on both Windows and Linux with proper dependencies installed. The QGIS plugin is currently only compatible with Linux.

Dependencies

Python 2.7.12

CUDA 6.0 or higher

Numpy

Boost.Python (Part of the Boost Libraries)

Python GDAL

PyCUDA

PyQt4 (Only needed for stand-alone GUI version)

This list may not include all of the dependencies the above programs require. Instructions on where to download dependedncies from and some installation tips are in Setup-instructions.txt.

The gpustruct class included here was taken from another git repo. Thanks to OSU!

Installation

Simply download this git repository to your local machine and put them wherever you want. If your CUDA installation folder isn't already set on your PATH, the plugin may not work properly. If you want to use the QGIS plugin with a GUI prompt, copy the folder "CUDARaster" to the QGIS plugin folder.

On Linux the default location is

/home/username/.qgis/python/plugins 

In Windows the default location is

 C:\Users\username\\.qgis\python\plugins

Use

This can be used via a command line interface or QGIS plugin.

Plugin: Has a simple GUI interface for chosing input and output files and calculating slope, aspect and hillshade. The plugin supports choosing between files on disk and layers already loaded into QGIS. Currently using active layers isn't completely accurate and can not be guaranteed to calculate everything properly.

Command line:

python scheduler.py input_file output_1 func_1 output_2 func_2 ... output_n func_n

About

Opensource GIS Tool leveraging NVIDIA CUDA and pyCuda for fast Raster Analysis

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published