Edge Enhancement

Overview

Edge Enhancement uses the image intensity gradient to enhance object boundaries. At any pixel, i, the result at that pixel is given by

    result(i) = input(i) * (fraction + (1 - fraction) * gradient(F(input))(i))

where fraction is a number between 0 and 1 and F is either a linear or an arctangent function.

The controls for selecting the input file, output file, and region to process are shared with many other Priism applications and are described elsewhere. The controls specific to Edge Enhancement are described below.

Topics

Region processing | Atan | Fraction | Min/max | Command line

Related Priism Topics

Priism

Atan

When this toggle is on, the arctangent of the input (after scaling it with the given minimum and maximum values) is used. When the toggle is off, the input is simply restricted by the minimum and maximum values.

In some cases, applying the arctangent function gives improved edge enhancement.

Return to overview

Fraction

This is a weighting factor between zero and one which determines the relative contributions of the original image and the gradient of the original. With a larger weighting factor, the result of the edge enhancement will look more like the input image.

Return to overview

Min/max

In the gradient calculation, the input image values are restricted to fall in the given range. Values outside the range are taken to be the value of the closest endpoint.

Return to overview

Command Line

Edge Enhancement accepts the command-line arguments described in Region.html. In addition, Edge Enhancement has the following options (optional parts are shown in brackets):

-atan

If present on the command line, this option causes the arctangent of the input data (after clamping to the range given by the minimum and maximum values) to be used. The default is to use the clamped input data without further modification.

-fraction=weight

Sets the weighting factor for combining the gradient of the input data and the input data. Valid values for weight are between zero and one with larger values giving a result that is closer to the input data.

-scl1=min[:max]

Sets the minimum and maximum used to clamp the first wavelength's data before calculating the gradient. The default is to use the minimum and maximum values recorded in the input's header. There are similar options (-scl2, -scl3, -scl4, and -scl5 to set the range for values drawn from other input wavelengths).

As an example, the following writes the gradient amplitude to gradient.dat for every other of the first thirty time points in series.dat; the arctangent is taken before computing the gradient:

    Edge series.dat gradient.dat -atan -t=0:29:2

Return to overview