|
Resolution
in a digital microscope system can refer to several very different things.
This is unfortunate because of the confusion the question of “what is
the resolution?” brings. First to consider is the resolution of the
camera itself. This is sometimes expressed as some number of lines, i.e.
800 lines resolution, or given as the number of pixels in the camera array,
i.e. 1600x1200. For line resolution the measurement is for the number of
black and white line pairs that could be resolved if the central, highest
resolution portion of the display were extended across the whole screen.
Pretty esoteric stuff since few people look at vertical black and white
lines. The number is, though, a relative indication of how sharp the image
will be - the higher the number the sharper the image that the camera is
capable of producing. The case with high resolution digital cameras is
almost straight forward since here all the pixels count- well most of them
do anyway. A camera with 1600x1200 sensor elements will make an image with
1600x1200 pixels. There is a little fudging because of the way the color is
derived (demosaicing), but that is another, also somewhat complicated
story. With the pixel shifted image however there is none of the
demosaicing and each pixel really sees a red and blue and green signal
– almost as if it were a 3 three chip camera. Now to the
microscope resolution part of the question. Because the microscope has a
zoom capability and can magnify an object to different sizes, the resolution
will be dependent on both the magnification and the resolving power of the
lens. The resolution of lenses is ultimately dependent on the wavelength of
light - in air you simply cannot see anything below about half the
wavelength of the illuminating light no matter how good the lens is or how
much it will magnify. Without getting too technical the theoretical limit
is about .33 micrometer laterally and about 1micometer vertically but the
practical limit is usually a little worse, perhaps closer to 1 micrometer
in the X/Y and 5 micrometers or more in the Z for lower magnification
lenses. In a perfect world, that would mean that you needed a minimum
magnification of roughly 300x to be able to just resolve two 1-micrometer
objects. In practical terms samples rarely have the extreme contrast or
other physical and optical properties necessary to allow that resolution at
that magnification. There is also the distinction that needs to be made
between resolving, which is more of a theoretical exercise, and in being
able to “see” the resolution. Sampling theory and optical
physics define the resolving power of the system but, in practical terms,
the magnification needs to be probably three to five times or more higher
than theory would predict in order to have enough information in the image
to actually “see” the resolution. Higher magnification is
indeed empty magnification but at least you will have the image spread
across enough pixels to allow you to visually distinguish between adjacent
objects.
|
