Signals and Biomedical Signal Processing
9
1.6 SOME CHARACTERISTICS OF DIGITAL IMAGES
Digital images (i.e., 2-D digital signals) are important types of data used in many
fields of science and technology. The importance of imaging systems (such as MRI)
in medical sciences cannot be overestimated. In this section, some general characteristics of images together with some simple operations for elementary analysis of
digital images are discussed.
1.6.1 IMAGE CAPTURING
Unlike photographic images in which cameras are used to capture the light intensity
and/or color of objects, each medical technology uses a different set of physical properties of living tissues to generate an image. For example, while MRI is based on the magnetic prosperities of a tissue, CT scan relies on the interaction between the x-ray beams
and the biological tissues to form an image. In other words, in medical imaging sensors of different physical properties of materials (including light intensity and color) are
employed to record anatomical and functional information about the tissue under study.
1.6.2 IMAGE REPRESENTATION
Even though different sensor technologies are used to generate biomedical images,
when it comes to the representation image, they are all visually represented as digital
images. These images are either gray-level images or color images. In a gray-level
image, the light intensity or brightness of an object shown at coordinates (x, y) of the
image is represented by a number called “gray level.” The higher the gray-level number, the brighter the image will be at the coordinate point (x, y). The maximum value
on the range of gray level represents a completely bright point, while a point with the
gray level of zero is a completely dark point. The gray points that are partially bright
and partially dark get a gray-level value that is between 0 and the maximum value of
brightness. The most popular ranges of gray level used in typical images are 0–255,
0–511, 0–1023, and so on. The gray levels are almost always set to be nonnegative
integer numbers (as opposed to real numbers). This saves a lot of digital storage space
(e.g., disk space) and expedites the processing of images significantly.
One can see that the wider the range of the gray level becomes, the better resolution is achieved. In order to see this more clearly, we present an example.
Example 1.1
Consider the image shown in Figure 1.4. Image (a) has the gray-level range of 0–255.
In order to see how the image resolution is affected by the gray-level range, we
reduce the range to smaller ranges. In order to generate the image with gray level
0–255, we divide every gray level of every point by two and round up the number
to the closest integer. As can be seen in image (b), which has only 64 levels in it,
the resolution of the image is not significantly affected by the gray-level reduction.
However, if we continue this process, the degradation in resolution and quality
becomes more visible (as shown in (c) which has only two levels of gray and dark
in it). Image (c) that allows only two gray levels (0 and 1) is called a binary image.