Histogram equalization alters the input data to give the closest possible approximation to data that is uniformly distributed over a given range. The alteration preserves the relationship between pixel values: a pixel whose input value is greater than another pixel's value will have its output value greater than or equal to the other pixel's output value.
HisEqual applies histogram equalization separately to each input section (so the preservation of the relationship between pixels only holds for pixels within the same section). The controls for the range of sections to process and the x and y bounds of the region to process are similar to those in other applications and are described elsewhere. The additional parameters that HisEqual uses are Range, which sets the input and output range of values (input values outside of the range are mapped to one of the endpoints of the range), and Ndiv, which controls how finely the range in order to calculate the approximation to the uniform distribution.
Overview | Region processing | Range | Ndiv
When HisEqual is applied to a section, it takes pixel values falling in a given range and maps them to the same range in the output; the mapping is nonlinear so that the output values are approximately uniformly distributed over the range. Input values below the range's minimum are mapped to the minimum; input values above the range's maximum are mapped to the maximum.
The first value in the Range field set the lower bound of the range; the second value is the upper bound.
Overview | Region processing | Range | Ndiv
To calculate the mapping between the input and output, the input and output ranges are divided into Ndiv equally sized bins. In general, using a larger amount of bins will improve how well the output approximates a uniform distribution though the discreteness and "clumpiness" (how many input pixels share the same value) of the input data are other limiting factors.
Overview | Region processing | Range | Ndiv