Convolution
The Convolution Extension
Define kernel with glConvolutionFilter2DEXT()
Allows per-component filters
May be faster
Probably will give more precise results
Most implementations use IEEE floating-point
Previous slide
Next slide
Back to first slide
View graphic version