I take a picture by microscope in which cells are present in different intensity zone. How I separate or segment them? For a particlur threshold it will show particular cell, changing the threshold how can I see the lower intensity zone cell and higher intensity zone cell and how I count them? I need a code for MATLAB.
When you do microscopy you should take a blank/background/flatfield image, meaning an image of a blank well or slide of the same type that your cells are in, using all exactly the same settings on the microscope. The flatfield image is basically the background that you want to remove the effects of. Then using Matlab, divide the cells image by the flatfield image. That should help to reduce the problems of uneven illumination etc. that is causing your segmentation problem.
If you really don't find it possible to take a flatfield image, say if you don't have access to that microscope any more, you could try to do it according to the example in the Matlab documentation on the imdivide function. They used a disk-shaped strel that's bigger than the grains of rice to remove them, then divided the original rice image by the imopen'ed image. So in your case you would have to use a strel that's bigger than your cells.
https://en.wikipedia.org/wiki/Flat-field_correction
http://fr.mathworks.com/help/images/ref/imdivide.html?s_tid=srchtitle
Hi Hirak,
Thanks very much for the example image. I also see that you have found and signed up for our new software MIPAR at http://MIPAR.us -- thanks very much!
I have used MIPAR to create a recipe file which finds your features quite nicely. I have attached this recipe file along with the segmentation it produces.
Please contact me with any questions or thoughts!
Cheers,
John
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing