How straight foward is it to create a large image montage using images acquired using a microscope using a motorized XYZ stage?
this depends on what kind of montage you wish to make, If you wish to make a normal merging its simple and can be done using Photoshop but if you want to do a 3D stacking, it might take time & specific microscope with softwares to form monatge
http://digital.bsd.uchicago.edu/resources_files/making%20montages%20with%20ImageJ.pdf
Thanks for the feedback Disha.
We have attempted to use the ImageJ plugin from the National Centre for Microscopy but have had a number of problems - particularly with the entire mosaic looking badly tiled. I think this can be dealt with using some form of background image correction procedure but I don't have sufficient experience to know how to do this.
If your images have reasonable overlap, you could give the Fiji Stitching plugin a try:
http://fiji.sc/wiki/index.php/Image_Stitching
Dear Stephan,
Thank you very much for your suggestion. To provide you with a little more detail I have acquired images using a motorized Olympus FLV1000 of one hemisphere of the rat brain. The Olympus machine has software that allows you to acquire a whole section at whatever magnification you desire, and also to collect each panel in the z-dimension. We have been using the montage application from The National Center for Microscopy in California. This does a reasonably good job of reassembling the images. However, what I'm not satisfied with is the procedure that is doing the background correction. It is my understanding that if each tile in a mosaic is not background corrected it results in the overall mosaic having a clearly defined tiled appearance. When background correction is applied this disappears and the image looks seamless. I note that Fiji does have an option for background correction but the challenge is that this needs to be done one image at a time and we have several thousand to do. I was wondering whether it might be possible to write a background subtraction routine in a program such as Matlab. However, I have had trouble in finding exactly what is being done when background subtraction routine is run. My guess is that it is effectively calculating the average luminosity across all the pixels in the background image and then proportionately increasing or decreasing the levels of every individual pixel to match the average. These values are then applied to the actual images which contain information. Does that sound right?
Have you tried the software "Imago"?
http://www.mayachitra.com/imago/
I've seen it used at Steven Fisher's Lab at UCSB where they have a similar setup to yours, i.e. Olympus FV1000 with motorized stage and controls, to image an entire mouse retina (~300+ z-stacks) and use the above software to stitch it all together.
Geoff Lewis from the Fisher Lab recently gave at talk at a meeting in Berlin and showed some stitched images and they looked pretty seamless. You might want to ask him for image samples or he can direct you to the engineers that develop imago in regards to what calculations they might be applying to the images.
http://www.nri.ucsb.edu/2people/lewis/lewis.html
- Badges
- Science method