Archived posting to the Leica Users Group, 2000/11/02

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: [Leica] Retry and missing link (2)
From: imx <imxputs@knoware.nl>
Date: Thu, 02 Nov 2000 17:39:12 +0100

When you digitize a picture (by scanning) or capture (CCD), we use a sensor
array of CCD elements of a certain dimension. Every individual sampled
(digitized) element is called a pixel. BUT: pixels are inherently without
physical dimensions! Here is the missing link. Image points as I used them
in the resolution calculation have a finite, but well defined diameter of
0.03mm and these points are physically located in the emulsion.
Not so in digital capture. A pixel is identified in an image file by its
spatial coordinates (rownumber/columnnumber) and the attributes of colour
and brightness. Jim Brick will surely confirm this. As it is so important I
will repeat: pixels have NO physical diamensions and have only a reference
in spatial coordinates. In other words: the size of the original sensor that
is used for sampling is not imcrporated into the data of the image file.
The file tells you that at columnlocation 15 and rowlocation 5, there is
associated a color and a brightness. This info is the same irrespective of
the size of the individual sensor (large or small). The digitized picture is
represented in the image file as a matrix of N by M dimensions (rows,
columns). Of course the finer the sensor grid (size of sensor elements) the
larger the matrix. But large or small, this matrix is without any physical
dimension. 
We simply have a list like this one:
location   colour   brightness
15,1        133         23
15,2        4           19
15,3        778         111

This list would be the same if our individual sensor elements are 3 micron
or 30 micron in diameter. Of coure a  3 micron element can capture more
detail than can a 30 micron sensor.
The simple notation: this CCD has 960 x 1440 picture elements (sensor
elements), gives you no clues about the size of individual sensors. It only
defines the size of the N x M matrix in the file.
We all know that the pixel array in de digial camera, needs more than one
sensor element to capture the necessary information about location,
brightness and colour. Generally we need 4 sensor elements to define one
cell in the matrix above. The CCD of 960 x 1440 sensor elements can capture
only information for a 480 x 720 matrix in the image file.
When do dimensionless pixels get their physical size or dimension?
See part 3