Catalogue of Artificial Intelligence Techniques


Jump to: Top | Entry | References | Comments

View Maths as: Images | MathML

Edge Detection

Keywords: resolution cones, zero-crossings

Categories: Pattern Recognition and Image Processing , Vision

Author(s): Bob Beattie

The purpose of edge detection is to locate places in images which correspond to informative scene events, such as a shadow or the obscuring boundary of an object. The projection of such events often produces steep gradients and discontinuities in image intensity, so the basic methods of edge detection are:

Computationally, this normally involves Convolution of a neighbourhood operator with the image at every point in the image. Very many operators have been implemented, but because of the difficulty of precisely specifying the goal of edge detection, it is difficult to compare operator performance. The latest theories of edge detection involve applying operators to the same image at different resolutions in resolution cones in an attempt to provide a better description of the intensity changes present in the image. See also Boundary Detection.



Add Comment

No comments.