Convolution Modules

Convolution is a word for "wrapping up something by something else". This term tries to describe how our brain perceives the environment by use of the eyes. First, our brain looks for certain basic forms (right angle, straight line, bow and so on). For that it uses fixed detectors that recognize a specific form and classify it, according to the position of one part in respect to the others. The brain convolutes the results to what we "see" in the end. It works massively parallel, i.e. all detectors are doing their work at the same time, and the sum of all detection results will be presented to our consciousness as visible objects.

A technical convolution reduces this total process to one single function, e.g. "recognize a vertical line", and shows the result in an intelligible form (graphical). GoDot's most powerful convulution module is mod.Convolve. All other listed modules represent special solutions and extensions to the basic algorithm.


Arndt Dettke

Copyright © 1997, A. Dettke, Last Updated - 10/16/97