Catalogue of Artificial Intelligence Techniques
Difference of Gaussians
Categories: Pattern Recognition and Image Processing
Author(s): T.P. Pridmore, S.R. Pollard, S.T. Stenton
This function is composed of the difference of two Gaussian distributions and approximates (the Laplacian of a Gaussian), the operator which Marr and Hildreth proposed to be optimal for Edge Detection in images. It also describes the `mexican hat' weighting function of the receptive fields of retinal ganglion and LGN cells.
- Marr, D. and Hildreth, E., Theory of edge detection, Proc. Roy. Soc. Lond., vol.B207, 1980, pp.187--217, also appears in Computer Vision: Principles,
(Kasturi, R.J. and Jain, R.C., eds.), Las Alamitos,
CA, IEEE Computer Society Press, 1991, pp. 77--107