On 7/3/07 4:00 PM, "IMAGEJ automatic digest system" <
[hidden email]>
Antje posed a series of questions around focus.
1. Focus measurement (more general image analysis question) (11)
Noel says,
1. Confocal images should be in focus, by definition (depending upon the
aperture size of course), if they are blurry, it means there has been
movement or maybe heat produced changes in refractive index (heatwaves) or
maybe vibration, or who knows. Maybe even the objects are degraded.
However, in-focus usually means sharp.
And as there is no stack of Z images we are restricted to a single focal
image, so we must look for some criteria which suits.
A differential of the image will indicate degrees of sharpness.
Try a sum based on the absolute values of the differences between the
central pixel and the surrounding 8. Add the differences all together
ignoring the sign. The bigger the sum the sharper the group of 9 pixel. (Of
course if you have noise then you will also get a big sum). Do this for
every pixel in the image and keep the sharpness result.
It would be the work of a few days to create a plug-in to run through a
series of images, and produce a set of sharpness images and sharpness
results, which could then be used to sort the originals. So long as the
images are not saturated black or white, the sum will vary with sharpness
and not be influenced by local intensity changes. Of course the camera
response curve may alter relationships.
Astronomers use software to pick out sharp from fuzzy in a series of images.
Another tack would be to use the FFT of the image and examine this to see
where the detail starts/ends. This would be a radius on the FFT image. The
smaller the radius at the transition, the higher the resolution (and hence
the sharpness). This would need another plug-in.
Or you could even try to do the convolution/deconvolution route, using the
point spread function for the microscope, and deconvolving that with the
images. This can recover data from out of focus regions, but may need z
series data to do well.
--
Noel Goldsmith
DSTO
506 Lorimer Street
Port Melbourne
Victoria
Australia
3207
Phone 613 96267527
Fax 613 96267718
Email
[hidden email]
IMPORTANT: This email remains the property of the Australian Defence Organisation and is subject to the jurisdiction of section 70 of the CRIMES ACT 1914. If you have received this email in error, you are requested to contact the sender and delete the email.
Locked
