Histogram Equalisation

Histogram Equalisation

One method of enhancing an image is commonly called "histogram equalisation". With this technique, the intensity histogram of an image is first calculated. From this an adjustment is calculated so that the resultant image has a cumulative histogram which is more nearly linear. The aim is to try to make more effective use of the overall intensity range available.

Equalise what buffer ?

A list of buffers may be specified to which the histogram equalisation is to be applied. Any valid buffer may be included.

Another buffer ? If a further buffer is to be included in the list, then yes (<Y>) should be specified. No <N> indicates that all the buffers have been specified.

Equalisation based on Window or whole Screen ?

The histogram is evaluated using either a window (<W>) or the entire contents of a buffer (<S>). The equalisation may subsequently be applied to either the entire buffer or a different window.

Window Selection: Region to base equalisation on

The window selection submenu is produced if the equalisation is to be based on a window. The required window is specified in the normal manner. Further information is available through the [H Help] option within the submenu.

Apply equalisation to Window or whole Screen ?

The modification calculated on the basis of the histogram may be applied to either a window (<W>) or the entire contents of a buffer (<S>). This window is independent of that through which the modification is calculated.

Window Selection: Region to apply equalisation to

The window selection submenu is produced if the equalisation is to be applied to a window. The required window is specified in the normal manner. Further information is available through the [H Help] option within the submenu.

Equalisation factor (1:full, 0.5:half, 0:none etc.) ?

The equalisation factor controls the degree to which the equalisation is applied. A value of 0 causes the image to remain unchanged, 1 full equalisation to be made. Negative numbers cause the intensity distribution to be less linear, while values greater than 1 cause over correction. Values lying between 0 and 1 give partial equalisation.

Lower cut off (do not equalise beyond this point) ?

Intensity values falling below this value will not be affected by the equalisation process, nor used in its calculation.

Upper cut off (do not equalise beyond this pont) ?

Intensity values falling above this value will not be affected by the equalisation process, nor used in its calculation.

Smooth mapping function ?

As an option, the mapping function relating the current with the modified intensities may be smoothed <Y>. This smoothing will not generally be required or have much effect, except if there are nearly discontinuities in the intensity information.

Parent menu

Main Menu

DigImage User Documentation


Stuart Dalziel, last page update: 19 February 1996