Due to the nature of the application, the software requires a reasonable amount of system memory. We recommend 4 GB of RAM or more for multispectral images and 8 GB of RAM or more for hyperspectral images. You also need a mid-range to high-end CPU. Note that the GPU (graphics processing unit) is of special importance due to heavy drawing operations, make sure not to use a low-end graphics card. Amongst others, middle-grade nVidia Quadro perform reasonably well, even their mobile versions.
The following libraries and tools are necessary to build and run Gerbil:
- CMake build system (version >= 2.8.11)
- boost C++ libraries, modules
filesystem(version >= 1.47)
- Intel TBB parallelization framework
- Qt application and UI framework (version >= 5.5.1) (download)
- OpenCV computer vision library (version >= 2.4)
- GDAL geospatial data abstraction library
If you are having trouble obtaining a recent version of Qt for your system, you might try the
which requires Qt version 4.8.