Related Experiment Video
Updated: Aug 26, 2026

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
The influence of edge detection algorithms on the estimation of the fractal dimension of binary digital images
Helmut Ahammer1, Trevor T J DeVaney
1Institute of Medical Physics and Biophysics, University of Graz, Harrachgasse 21, A-8010 Graz, Austria. helmut.ahammer@uni-graz.at
Abstract:
The boundary of a fractal object, represented in a two-dimensional space, is theoretically a line with an infinitely small width. In digital images this boundary or contour is limited to the pixel resolution of the image and the width of the line commonly depends on the edge detection algorithm used. The Minkowski dimension was evaluated by using three different edge detection algorithms (Sobel, Roberts, and Laplace operator). These three operators were investigated because they are very widely used and because their edge detection result is very distinct concerning the line width. Very common fractals (Sierpinski carpet and Koch islands) were investigated as well as the binary images from a cancer invasion assay taken with a confocal laser scanning microscope. The fractal dimension is directly proportional to the width of the contour line and the fact, that in practice very often the investigated objects are fractals only within a limited resolution range is considered too.
Related Concept Videos
Downsampling
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
Discrete Fourier Transform

