Visual Prosthesis Based on Optic Nerve Stimulation with Penetrating Electrode Array - Artificial Sight: Basic Research, Biomedical Engineering, and Clinical Advances - page 199

Histogram Equalization

Histogram equalization is one of the popular image enhancement transformations.
It is used to obtain a uniform histogram and can improve contrast of the image.
This algorithm can be used on a whole image or just a part of the image.
Histogram modeling [13] is usually introduced using continuous, rather than
discrete, process functions. Therefore, we suppose that the image of interest
contains continuous intensity levels (in the interval [0,1]) and that the trans-
formation function f which maps an input image A ( x,y ) into an output image
B ( x,y ) is continuous within the interval. Furthermore, we assume that the
transfer law (which may also be written in terms of intensity density levels,
e.g. H B ( s ) f ( H A ( r ))) is single-valued and monotonically increasing (as the

case in histogram equalization) so that it is possible to define the inverse law
H A ( r )

=

f − 1 ( H B ( s )). An example of such a transfer function is illustrated in
Figure 10.11.

=

Edge Detection

The edges of an image hold much information, such as position, texture, shape
and size. An edge is where the intensity of an image moves from a low value to
a high value or vice versa.
The human visual system is very sensitive to details, especially abrupt changes
or edges. These details in the frequency domain are always located in the high
frequencies.
One way to detect edges or variations within a region of an image is by using
the gradient operator. For instance, the gradient G is a vector with two elements
G x and G y , where G x is the gradient in the width direction and G y is the gradient
in the height direction. Since G is a vector, its magnitude G m and direction angle
can be given as:

s s

255

ds

H B (r)

r

0 0

H A (r)

r

dr 0 255

Figure 10.11. Example of histogram transformation function.