SLIC_CUDA

How to run:

Ensure "rai" binary is in parent directory (same directory as rai_SLIC_CUDA). (Or modify run script) ./run.sh Observe output and build folder for results. ./del.sh to delete previous run and get ready for the next

Superpixel Segmentation SLIC of Achanta et al. [PAMI 2012, vol. 34, num. 11, pp. 2274-2282]. GPU version with CUDA.

Library required :

• opencv 3.0 min
• cuda compute capability 3.0 min

Example

Cuda : 30 fps

Comparison with cpu version from OpenCv : 4 fps