Tweet
Date: Monday, November 30, 1992
To: Ed Grood, interested Biomch-L readers
From: Joe Sommer, Penn State/NIH
Subj: cell alignment image analysis
Ref: Biomch-L posting 11/11/92
"Response of fibroblasts to mechanical deformation"
1) While a 2D FFT can provide overall image angular orientation
information, I do not believe that it can provide angular
DISTRIBUTION information such as you desire. This is based on
personal experience in 2D FFT image analysis of quarry faces to
assess rock fracture spacing/orientation for mining safety, and
of polyurethane foam pore size/orientation for manufacturing
quality control.
2) If the cells are elongated, a linear Hough transform may be
able to provide orientation distribution information. The Hough
transform is a histogram method that characterizes straight lines
in a 2D image by their slope angle THETA and their common
perpendicular from the origin RHO, and produces a new image in
the (RHO,THETA) domain where pixel intensity codes length of
individual line segments with that slope slope and offset.
Collapsing the 2D Hough image along the RHO axis after suitable
thresholding, may be able to give an idea of the orientation
distribution.
3) Without seeing a sample image, I am hesitant to recommend
automated image enhancement/segmentation methods "to separate
overlapping cells, remove dirt objects, etc.", in that my
experience has shown these tasks to be highly application
specific. If you can prepare a sample image for Internet
transfer, I would be happy to correspond further and perhaps even
run a sample 2D FFT or Hough for you.
Best wishes,
Joe Sommer
H.J. Sommer III, Professor of Mechanical Engineering, 207 Reber Building
The Pennsylvania State University, University Park, PA 16802
(814)865-2519, FAX (814)863-4848, Bitnet HJS@PSUECL, Internet HJS@ECL.PSU.EDU
SABBATICAL ADDRESS (8/15/92-6/15/93)
H.J. Sommer III, Guest Researcher, Biomechanics Lab, Rehabilitation Medicine
National Institutes of Health, Bldg 10, Room 6s-207a, Bethesda, MD 20892
(301)496-9890x15, FAX (301)402-0663, Internet SOMMER%BMLVAX.DNET@DXI.NIH.GOV