Subject: | |
From: | |
Reply To: | Martin W. Wessendorf |
Date: | Thu, 16 Feb 1995 12:27:34 CST |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
We're experiencing a problem on our MRC 1000 and I need help from other users.
In our images, rows and/or columns of pixels are periodically darker. As you
proceed across along a horizontal line, alternating pixels are darkened. In
addition, as you proceed down along a vertical line, roughly every 4th or 5th
horizontal line of pixels is darkened. Neither seem to change with scan speed.
Moreover, the placement of the lines--including the horizontal lines--stays the
same. (No, they aren't the lines on the monitor--you can rotate the image and
the lines rotate as well).
These lines initially aren't obvious, although they jump out if you run a 3x3
local contrast enhance filter (using comos) or a sharpening filter ("sharpen
more" using Photoshop). They are particularly problematic along the edges of
objects; this may reflect their showing up mostly in particular ranges of gray
level. Although not obvious, they can create problems in surface
reconstructions of cells--e.g., they can make a cell's contour look as if it
were cut out with pinking shears.
The periodicity of the horizontal lines is roughly what you'd expect from 120 Hz
power supply noise; we can see such noise in the power output of the laser, but
it's unclear how such noise could be synchronized to the beginning of the
scan--which it appears to be.
Has anyone else seen this? More to the point, does anyone know how to fix this?
Thanks--
Martin Wessendorf
|
|
|