Image Processing Reference
In-Depth Information
FIGURE 2.41
LENA image at a resolution of 3 of the original image.
2.8 IMAGE ENHANCEMENT
The goal of image enhancement is to provide a more instinctively pleasing image.
Enhancements are also used to simplify visual interpretation and understanding of
images. There are different techniques used in image enhancement such as unsharp
masking and histogram equalization [5,19
21].
-
2.8.1 U NSHARP M ASKING
The operation of unsharp masking is used to sharpen an image by enhancing its high-
frequency components. The process consists of the following three steps.
is slightly blurred to create a smooth (low-pass) image f ( n, m )
1. Image f ( n, m )
.
2. Difference between the original image and the smooth image is d ( n, m ) ¼
f ( n, m ) f ( n, m )
. This image consists mainly of high frequencies of the
original image.
3. Difference image is boosted by a factor of
and added back to the original
image to create a sharper image. The degree of sharpness is controlled by
the boost factor and the size of the LPF
b
f UM ( n, m ) ¼ f ( n, m ) þb( f ( n, m ) f ( n, m ))
(
2
:
121
)
The block diagram representing the unsharp masking process is shown in Figure 2.42.
In digital processing, the entire process of linear unsharp masking can be compactly
represented as a single high-pass FIR
filter operation.
Consider performing unsharp masking using an FIR
filter h LP ( n, m )
as the
blurring operator and a boost factor of
b
. The
filter used in this process is denoted
by
d( n, m )
and has all components as zeros except for the center tap (which is one),
Search WWH ::




Custom Search