Digital imaging Analogue and digital images
Historically, radiological images were recorded on glass plates covered with a light-sensitive emulsion. This emulsion, sensitive to a wide range of photon energies, consisted of microscopic particles of silver compounds which following exposure, developing and fixation built up an image, consisting of small black dots, of the radiographed object. Different types of intensifying screens and filmbased materials have not changed the basic principle that the information content of the image consists of microscopic dark dots, the dens it y of which determines the darkness in a given image area. The darkness thus varies continuously and the image produced is called an analogue image.
If the darkness along a line across an analogue image is measured with a densitometer as in Fig. 1 a, the result will be an irregular curve where the finest irregularities represent the individual silver particles. If the registered curve is divided into equal parts the dens it y in each part can be calculated as an average that can be given a numerical value. The averages are seen in Fig. l b and the corresponding numerical values in Fig. 1 c. When the whole surface of a radiological film is measured with a densitometer line by line in this way the analogue image can be transformed into a so called digital image (Fig. 2).
The distances between the lines and the size of the equal parts dividing each line defines the resolution of the digital image. A quadrilateral, as high as the distance between two lines and with the width of one division along a line, is called a picture element or a pixel.
The digital image is naturally adapted to computer techniques. Thus, the images are normally divided into a number of pixels equal to powers of 2, e.g. 512 x 512, 1024 x 1024 or 2048 x 2048 pixels. The
|
|
Figure 1. Analogue to digital conversion of a photometry recording along a line into a sequence of binary Y signals.
A) Schematic photometry curve of a line, e.g. across a radiographic film
B) The same curve divided into equal distances. For each segment the average photometer recordings have been assessed on a scale from 0 to 15 (24 steps)
C) Digital printout of the photometer averages in 1 B
D) Binary representation of the digital series of numbers in 1 C.
|
Figure 2.
The influence of spatial and density resolution on image quality in an analogue to digital image transformation.
a) Analogue image with moderate spatial and density resolution. A low spatial resolution matrix is superimposed upon the image.
b) The same analogue image with a 10 by 10 pixel spatial matrix and 2 bits contrast resolution, i.e. 4 density levels.
e) The final digital representation of the low spatial and density resolution image. |
number of possible shades between black and white are also often referred to by a binary value, for example 8 bits = 28 or 256 shades, 10 bits = 1024 shades, or 12 bits = 4096 shades. In Fig. l d the darkness is given as 4 bits or 16 shades or levels and in Fig. 2 as 2 bits or only 4 levels.
Tatsuo Kumazaki and Hans Ringertz