Robert Hero and Alex Pang
06/30/2007 09:00 AM
Computer Science
This paper presents hardware acceleration techniques for both the LIC and HyperLIC algorithms using off-the shelf graphics hardware. The methods used in this paper take the burden of calculation from the CPU and place it on the graphics card through the use of programmable fragment shaders and texture maps. Frame rates increased almost 2 orders of magnitude over the corresponding software implementations of Fast LIC and HyperLIC algorithms. The methods presented here can also be extended to 3D with similar performance gains.
