gaussian sigma variance value
width of the kernel matrix
height of the kernel matrix
t { import std.algorithm.comparison : equal; import std.math : approxEqual; auto log = laplacianOfGaussian!float(0.84f, 3, 3); auto expected = [0.147722, -0.00865228, 0.147722, -0.00865228, -0.556277, -0.00865228, 0.147722, -0.00865228, 0.147722].sliced(3, 3); assert(equal!approxEqual(log.flattened, expected.flattened)
Create laplacian of gaussian $(LINK3 http://homepages.inf.ed.ac.uk/rbf/HIPR2/log.htm, (LoG)) filter kernel.