EM 1110-2-2907
1 October 2003
it is possible that a number of the pixels have DNs greater than 256. An image en-
hancement will decompress these values, thereby increasing their contrast.
Data Analysis
Histograms
Image processing software can chart the distribution of digital number
values within a scene. The distribution of the brightness values is
displayed as a histogram chart. The horizontal axis shows the spread of
the digital numbers from 0 to the maximum DN value in the data set.
The vertical axis shows the frequency or how many pixels in the scene
each value has (Figure 5-7). The histogram allows an analyst to quickly
access the type of distribution maintained by the data. Types of
distribution may be normal, bimodal, or skewed (Figure 5-7).
Histograms are particularly useful when images are enhanced.
Lookup Tables
A lookup table (LUT) graphs the intensity of the input pixel value
relative to the output brightness observed on the screen. The curve does
not provide information about the frequency of brightness, instead it
provides information regarding the range associated with the brightness
levels. An image enhancement can be modeled on a lookup table to
better evaluate the relationship between the unaltered raw data and the
adjusted display data.
Scatter plots
The correlation between bands can be seen in scatter plots generated by
the software. The scatter plots graph the digital number value of one
band relative to another (Figure 5-8). Bands that are highly correlated
will produce plots with a linear relationship and little deviation from
the line. Bands that are not well correlated will lack a linear
relationship. Digital number values will cluster or span the chart
randomly. Scatter plots allow for a quick assessment of the usefulness
of particular band combinations.
5-14
Previous Page
