Page 297 - Fundamentals of Light Microscopy and Electronic Imaging

280 DIGITAL CCD MICROSCOPY

• Indicate the width and height of a full-frame image (in pixels). Calculate the
image size of a print made using a 1200 dpi printer.

Camera Sensitivity and Noise
• What is the range of wavelength sensitivity? Does the CCD contain an
enhanced coating for UV or IR wavelengths? What is the QE at 550 nm?
• What is the saturation value (full well capacity) of a pixel in electrons?
• What is the read noise per pixel (in electrons) of the camera? Determine the
mean value of a bias frame to establish the contribution from bias plus read
noise.
• Calculate the percent contribution of the bias signal in an image signal that
reaches saturation.
• What is the operating temperature of the CCD and what is the contribution of
thermal noise? For quantitative applications, thermal noise (and read noise)
should be minimal. The noise levels are less important when acquiring
images to prepare prints, especially if specimens are bright.

Dynamic Range
• Calculate the dynamic range (maximum number of gray levels) from the full
well saturation level and the camera read noise (both values must be in units
of electrons).
• Note the bit depth of the digitizer (8, 10, 12, 14 bits). This value is sometimes
assigned to the camera as a whole.
• Is the bit depth of the digitizer a good match for the camera’s dynamic range?
The fact that the digitizer may be oversized does not condemn a camera, but
you should be aware of (and calculate!) the compensating gain factor applied
by the company to allow the camera to use all of the gray-level values. This
factor is often overlooked, but it is important when evaluating and comparing
the performance of different cameras. Remember that high gain settings
(fewer electrons/ADU) give grainier images.

Temporal Resolution
• What is the minimum time required for the camera to acquire and display a
full-frame, unbinned image? A one-quarter frame binned 2 2? To do this,
use a stopwatch and adjust the software for the acquisition of 10 frames in
time-lapse mode, with a minimum time interval (1 ms) between the frames.

Image quality (qualitative and quantitative aspects)
• Histogram stretch a feature spanning a limited number of pixel values and
examine the quality of the displayed image in terms of noise and graininess.
• After histogram adjustment, print the image and examine its quality.
• Prepare a plot showing the ADUs of a row of pixels through an object. Exam-
ine the pixel fluctuations across a uniform feature and confirm that they are
minimal.