Centre-Surround cell (DoG) operator and Dot-pattern selective cell operator

N. Petkov and W.T.Visser, University of Groningen,
Department of Computing Science, Intelligent Systems

Other on-line image processing and computer vision applications are available at http://matlabserver.cs.rug.nl. You can download the source of this web application here.

This site offers a program that implements the physiologically inspired model of the dot-pattern selective cell. The initial stage of this model generates individual dot candidates by a 2D surround Gaussian convolution, normalised on contrast. Based on these dot candidates, dots corresponding to a specific size are detected by an AND-type combination of lateral inhibition and winner-takes-all competition. A spot group detecting unit then detects gratings of these dots. Finally, the density of these pattern is computed by a weighted spatial summation.

You can select an image or upload your own image. To run the program you need to set the parameters relevant for the selected method and submit the job for execution. After the program completes (in a few seconds to a few minutes, depending on the size of the problem and the current system load) the input image and the resulting output image will be displayed in the page. You can save the results on your system.

Click here for explanations, references and examples.


More on-line image processing and computer vision applications: http://matlabserver.cs.rug.nl (opens new window)

Please wait while the image list is being populated...