Dear all.
Is there a plugin or macro for making the elements of to images overlap ( elements in image 2 being roughly at the same coordinates as in image 1)? ( I have searched online but could not find any) The procedure i was thinking of is to load the two images, ask the user to select to 2 points ( forming a vector) on the first image, then mark the two corresponding points on the second image. Then based on these marked points do a translate and a rotate to make the images overlap (i.e. Transform the second image so the two vectors overlap) . ( More points can be used to improve accuracy). The reason: I am working for a research group that has a protocol for analyzing muscle fibers. This protocol requires overlapping fibers from two slides with different stainings. Previously this used be done by loading the master slide in the microscope, capturing images of the areas of interest. Then load the second slide in the microscope, navigate to and capture corresponding areas of the second slide. In order to make the fibers of the different slides overlap the user needed to rotate the camera or stage since its impossible to place the fibers on the slides in a identical angle. Now we have now a fully automated microscope witch supports automated tile-scaning but because of the motorized stage and tile-scaning function lacks the ability to easily rotate the slides. This however presents the opportunity to do tile-scaning and then using software to align the fibers of the two slides (i am aware that doing the whole slide might run into memory constraints and accuracy problems, but even for a 2x2 tile-scan this approach should prove to be a good timesaver, and roughly as accurate). Any help or opinions is appreciated, With best regards, Tomas |
Hi Tomas,
using Fiji, you have at least two options: 1. Open both images and mark corresponding points in both using the point selection tool (index number must match) and use the plugin: http://pacific.mpi-cbg.de/wiki/index.php/Landmark_Correspondences to map one image into the other using a transformation model of your choice (translation + rotation is rigid). 2. Import the slides into TrakEM2 and use either of the manual alignment/transformation modes. That is less trivial in terms of the learning the program and less easy to script but provides a much nicer user interface (interactive overlay supporting various overlay modes). The underlying engine is the same: http://www.ini.uzh.ch/~acardona/trakem2_manual.html#transform_patch http://www.ini.uzh.ch/~acardona/trakem2_manual.html#landmarks Good luck and best regards, Stephan On Sun, 2011-06-12 at 21:59 +0200, Tomas Karlsson wrote: > Dear all. > > Is there a plugin or macro for making the elements of to images overlap ( elements in image 2 being roughly at the same coordinates as in image 1)? ( I have searched online but could not find any) > > The procedure i was thinking of is to load the two images, ask the user to select to 2 points ( forming a vector) on the first image, then mark the two corresponding points on the second image. Then based on these marked points do a translate and a rotate to make the images overlap (i.e. Transform the second image so the two vectors overlap) . ( More points can be used to improve accuracy). > > The reason: > > I am working for a research group that has a protocol for analyzing muscle fibers. This protocol requires overlapping fibers from two slides with different stainings. Previously this used be done by loading the master slide in the microscope, capturing images of the areas of interest. Then load the second slide in the microscope, navigate to and capture corresponding areas of the second slide. In order to make the fibers of the different slides overlap the user needed to rotate the camera or stage since its impossible to place the fibers on the slides in a identical angle. > > Now we have now a fully automated microscope witch supports automated tile-scaning but because of the motorized stage and tile-scaning function lacks the ability to easily rotate the slides. This however presents the opportunity to do tile-scaning and then using software to align the fibers of the two slides (i am aware that doing the whole slide might run into memory constraints and accuracy problems, but even for a 2x2 tile-scan this approach should prove to be a good timesaver, and roughly as accurate). > > Any help or opinions is appreciated, With best regards, > > Tomas |
It should also be possible to use the "Turbo Reg" plugin in conventional
ImageJ. You can search for it on the plugins page. Joel On Sun, Jun 12, 2011 at 5:45 PM, Stephan Saalfeld <[hidden email]>wrote: > Hi Tomas, > > using Fiji, you have at least two options: > > 1. > Open both images and mark corresponding points in both using the point > selection tool (index number must match) and use the plugin: > > http://pacific.mpi-cbg.de/wiki/index.php/Landmark_Correspondences > > to map one image into the other using a transformation model of your > choice (translation + rotation is rigid). > > > 2. > Import the slides into TrakEM2 and use either of the manual > alignment/transformation modes. That is less trivial in terms of the > learning the program and less easy to script but provides a much nicer > user interface (interactive overlay supporting various overlay modes). > The underlying engine is the same: > > http://www.ini.uzh.ch/~acardona/trakem2_manual.html#transform_patch > > http://www.ini.uzh.ch/~acardona/trakem2_manual.html#landmarks > > Good luck and best regards, > Stephan > > > > > On Sun, 2011-06-12 at 21:59 +0200, Tomas Karlsson wrote: > > Dear all. > > > > Is there a plugin or macro for making the elements of to images overlap > ( elements in image 2 being roughly at the same coordinates as in image 1)? > ( I have searched online but could not find any) > > > > The procedure i was thinking of is to load the two images, ask the user > to select to 2 points ( forming a vector) on the first image, then mark the > two corresponding points on the second image. Then based on these marked > points do a translate and a rotate to make the images overlap (i.e. > Transform the second image so the two vectors overlap) . ( More points can > be used to improve accuracy). > > > > The reason: > > > > I am working for a research group that has a protocol for analyzing > muscle fibers. This protocol requires overlapping fibers from two slides > with different stainings. Previously this used be done by loading the master > slide in the microscope, capturing images of the areas of interest. Then > load the second slide in the microscope, navigate to and capture > corresponding areas of the second slide. In order to make the fibers of the > different slides overlap the user needed to rotate the camera or stage since > its impossible to place the fibers on the slides in a identical angle. > > > > Now we have now a fully automated microscope witch supports automated > tile-scaning but because of the motorized stage and tile-scaning function > lacks the ability to easily rotate the slides. This however presents the > opportunity to do tile-scaning and then using software to align the fibers > of the two slides (i am aware that doing the whole slide might run into > memory constraints and accuracy problems, but even for a 2x2 tile-scan this > approach should prove to be a good timesaver, and roughly as accurate). > > > > Any help or opinions is appreciated, With best regards, > > > > Tomas > -- Joel B. Sheffield, Ph.D Department of Biology Temple University Philadelphia, PA 19122 Voice: 215 204 8839 e-mail: [hidden email] URL: http://astro.temple.edu/~jbs |
Free forum by Nabble | Edit this page |