User guide

Viewing and Editing Images
12-33
Step I: Identify Noise Characteristics
The first step in the Wizard is to identify the type of noise in your image.
Examine both the image and the density distribution histogram, then select
one, both, or neither of the following checkboxes:
• Salt. This type of noise appears as specks that are lighter than the
surrounding background. The density distribution histogram of this type
of noise displays noise peaks at the high end of the range (right end of the
plot). This type of noise is common in electronic cameras with
malfunctioning pixels. It can also be caused by dust or lint in the imaging
optics or scratches on photographic film. Salt is a type of outlier noise (see
below).
• Pepper. This type of noise appears as specks that are darker than the
surrounding background. The distribution histogram of this type of noise
displays noise peaks at the low end of the range (left end of the plot). Its
causes are similar to those of salt noise. Pepper is a type of outlier noise
(see below).
Next, select one of the following option buttons to describe additional
features of your noise.
• Gaussian. The distribution histogram of this type of noise has a Gaussian
profile, usually at the bottom of the data range. This type of noise is
usually an electronic artifact created by cameras and sensors, or by a
combination of independent unknown noise sources.
• Uniform noise. This type of noise appears in the histogram as a uniform
layer of noise across the data range of the image.
• Outlier noise. This category of noise includes salt and pepper noise (see
above). The distribution histogram of this type of noise displays noise
peaks at the high and low ends of the range.
After you have identified the type of noise, go to Step 2.
Step 2: Select Filter Size
Image noise is filtered by means of a filtering window (or kernel), which is
measured in pixels. This filtering window slides across the image, processing
the pixels within it.