Enhancement
Image enhancement adjusts pixel values to improve the perception of information in images for the human visual system. Enhancement techniques are not very useful as pre-processing operations for segmentation, except background subtraction (that is not directly an enhancement technique, but more a correction operation), which attempts to make uniform the illumination of the background to facilitate objects segmentation.

 

Manual Enhancement

Contrast: When increasing contrast, pixels that are darker than 127 (pixel's value) are made darker and pixels that are lighter than 127 are made lighter. This works fine when the values in the histogram is distributed around the middle; otherwise, contrast will not have the desired effect.

Luminosity: Increasing or decreasing the luminosity shifts the histogram.

Gamma: Moving the slider to the right (greater than one) will increase the number of dark values and darken the image; moving to the left (less than one) will increase the number of light values, lightening the image. The mapping function is an exponential curve.

Automatic Levels

Automatic levels (histogram normalization) stretch an image's pixel values to cover the entire pixel value range (0-255).

Equalized Levels

This operation attempts to flatten the image histogram. Histogram equalization employs a non-linear mapping which re-assigns the intensity values of pixels in the input image so that the output image contains a uniform distribution of intensities.

Background Subtraction

This operation attempts to make uniform the background illumination when, for instance, an image has been acquired by a microscope with a nonuniform illumination. There are two ways to correct this issue:

  1. Take a picture of the scene without specimen (only the background). Then, take another picture with the specimen.
    Select the image with the specimen in Source 1 and the one without the specimen in Source 2.
    Finally, adjust the Luminosity offset slider to correct the luminosity of the resulting image.
  2. When you do not have a picture with only the background, press the Create Artificial Background button.
    Set at least 6 points that match to the background of the image (using the Add Point button or Ctrl +W). The values of these points are used by a polynomial fit function to create an artificial background.
    Finally, press Add image in the Artificial Background dialog. Close the dialog and then follow the instructions at the point 1.

See also:

Image Factory