In this repo the parallel version of Mean Shift algorithm is presented; this repo belongs to the Master Degree Parallel Computing course project at UNIFI; here the link to its sequential version. Here the related [paper](https://github.com/pisalore/MeanShift_CUDA/blob/master.
For parallel Mean Shift algoithm implementation I used CUDA with Windows 10. In particular:
I strictly reccomend to read the CUDA installation Guide. A very common problem is linked to the VS2019 Linker using CUDA, and dependencies: under Project Properties > Linker > Input > Additional Dependencies remember to copy: cudart_static.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies) in order to correctly link cuda library. Debug has to be x64 based.
