Image Processing Reference

In-Depth Information

Fig. 5.17
(
a
) The principal behind image sharpening. (
b
) An example of image sharpening with

c

0
.
6. The pixel values of a horizontal line (the location is indicated by the
white line
in the
top

image
)areshowntothe
right

=

Eq.
5.9
. That is, we first calculate the absolute value of each pixel in the two images

and then add them together. The result is the final edge enhanced image. After this,

the final task is often to binarize the image, so that edges are white and the rest is

black. This is done by a standard thresholding algorithm. In Fig.
5.16
the final edge

enhanced image is shown together with binary edge images obtained using different

thresholds. The choice of threshold depends on the application.

5.2.3 Image Sharpening

The method presented in Sect. 4.4.2 and illustrated in Fig. 4.23 is not only applicable

to thresholding, but can also be used to increase the overall contrast of the image.

The method is expressed as follows in terms of correlation:

−
f(x,y)

h(x,y)

g(x,y)

=

f(x,y)

◦

(5.12)

where
h(x,y)
is a large mean filter kernel. The method belongs to the class of meth-

ods aimed at sharpening or enhancing the image. Another method for this purpose

is based on image edges and is explained in the following.

What makes it possible to see an object in a scene, is the fact that the object is

different from the background. From this follows that the transition between object

and background is of great importance and this is of course exactly why we measure