Web-enabled image processing operators

Below you find web implementations of various biologically motivated image processing operators, which you can use with your own image material.

• Gabor filter for image processing and computer vision. [download source]
  With this program you can (see explanations and examples):
  • visualize Gabor functions,
  • use a Gabor filter and a Gabor-energy filter,
  • perform edge detection,
  • simulate simple and complex cells (visual cortex),
  • extract texture features,
  • simulate non-classical receptive field inhibition or surround suppression (visual cortex),
  • perform contour detection,
  • explain certain visual perception effects.

• Canny edge detector (Canny filter) for image processing and computer vision. [download source]
  With this program you can (see explanations and examples):
  • apply the Canny edge detector on-line on your own images,
  • visualize a derivative of a Gaussian function,
  • simulate non-classical receptive field inhibition or surround suppression,
  • perform contour detection.

• Centre-Surround cell (DoG) operator and Dot-pattern selective cell operator. [download source]
  This program can be used to (see explanations and examples):
  • visualize Difference of Gaussian (DoG) functions,
  • simulate retinal-ganglion and LGN cells,
  • compute the convolution of an input image with a DoG (center-surround cell responses),
  • detect intensity spots,
  • detect groups of intensity spots,
  • detect dot-pattern texture.

• Grating cell operator for image processing and computer vision. [download source]
  This program can be used to (see explanations and examples):
  • Detect grating
  • Visualization of grating subunits
  • Visualization of grating cells

• CORF operator for contour detection
  The CORF operator shares similar properties with simple cells in visual cortex. It uses center-on and center-off Difference of Gaussians (DoG) functions, which model LGN cells, and combine their responses with a weighted geometric mean. Besides orientation selectivity, it exhibits cross orientation suppression, contrast invariant orientation tuning and response saturation. You are kindly invited to use this matlab script referring to the following article:
  

  G. Azzopardi and N. Petkov: "A CORF computational model that relies on LGN input outperforms the Gabor function model", Biological Cybernetics(1), 2012, DOI: 10.1007/s00422-012-0486-6. [pdf],[bib]

  This script takes as input:
  • I - Intensity image
  • sigma - the outer standard deviation of the DoG Gaussian function. The standard deviation of the inner DoG Gaussian function is 0.5 of sigma
  • t - The high threshold that is used for hysteresis thresholding. The low threshold is set to half of the given high threshold
  It gives the following output:
  • A grayscale image where a pixel value, which varies between 0 and 1, corresponds to the strength of the presence of an edge at the concerned position
  • A binarized contour map based on the given high threshold t