Hi,
I analyze invertebrate diversity on bottom-ocean images and I find many invertebrate species on each photo. However, I can only add up to 38 counter types in the Cell Counter plugin since the number of the counter types that fit on the screen depends on the size of my computer screen (!?) from the top to the bottom of the screen. You can imagine that I am using the biggest screen I can find. I am wandering whether anyone ran into this problem and maybe solved it by fixing the plugin so that you can scroll up and down the cell counter window, once you ran into the bottom of the screen. I really probably need around 200 counter types, since I have several stations and the species only partially overlap between the stations. So far, I have been assigning Types 1-38 for different species at each station but this eventually leads to a major confusion and a lot of additional work on data editing. Thank you, Pavica |
Hi Pavica,
I don’t have a direct solution to your problem, but it reminds me one concern I had a few weeks ago (and that I overcame by a different way) and thus I take advantage of your thread to make a suggestion of a new feature to the CellCounter developper… Will be doable to have editable name for each counter ? Using a text field (maybe limited in size). If this configuration could be save and load (in text/csv file), it will be even more usefull. Cheers, Romain --------------------------------------------------------------- Dr. Romain Guiet Bioimaging and Optics Platform (PT-BIOP) Ecole Polytechnique Fédérale de Lausanne (EPFL) Faculty of Life Sciences Station 19, AI 0140 CH-1015 Lausanne Phone: [+4121 69] 39629 http://biop.epfl.ch/ --------------------------------------------------------------- ________________________________________ De : ImageJ Interest Group [[hidden email]] de la part de Pavica [[hidden email]] Envoyé : vendredi 27 février 2015 08:29 À : [hidden email] Objet : How to add more than 38 counter types in Cell Counter plugin Hi, I analyze invertebrate diversity on bottom-ocean images and I find many invertebrate species on each photo. However, I can only add up to 38 counter types in the Cell Counter plugin since the number of the counter types that fit on the screen depends on the size of my computer screen (!?) from the top to the bottom of the screen. You can imagine that I am using the biggest screen I can find. I am wandering whether anyone ran into this problem and maybe solved it by fixing the plugin so that you can scroll up and down the cell counter window, once you ran into the bottom of the screen. I really probably need around 200 counter types, since I have several stations and the species only partially overlap between the stations. So far, I have been assigning Types 1-38 for different species at each station but this eventually leads to a major confusion and a lot of additional work on data editing. Thank you, Pavica -- View this message in context: http://imagej.1557.x6.nabble.com/How-to-add-more-than-38-counter-types-in-Cell-Counter-plugin-tp5011798.html Sent from the ImageJ mailing list archive at Nabble.com. -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Hi Pavica & Romain,
> Will be doable to have editable name for each counter ? I added the Cell Counter plugin to the Fiji distribution of ImageJ [1]. Find it in the Plugins > Analyze menu (or better: press L to easily launch plugins). I forked the code from the version on the ImageJ1 web site, cleaned up and updated the code. I added a simple Rename feature for renaming the counter types. Unfortunately, I did not have time to make it save the new labels when saving XML etc. I also investigated adding a scroll bar to the counter types (you'd think it would be easy), but was not able to do it with reasonable effort because the code uses a complex "grid bag" layout. It would take at least some hours to migrate the code away from that type of layout, so that a scroll bar could easily be added. At least now the code is on GitHub [2], so that others can easily contribute such enhancements if they have time. Finally, I would strongly urge those using this plugin to seek a more automated way of counting your structures, especially if you need more than 38 (!!!) counter types. You really want to find a way to make your published results reproducible, in the name of good science [3]. Regards, Curtis P.S. to the original author (Kurt De Vos): let me know if you want to be added as an admin on the GitHub repository. [1] http://fiji.sc/Cell_Counter [2] https://github.com/fiji/Cell_Counter [3] http://loci.wisc.edu/software/oss On Fri, Feb 27, 2015 at 2:19 AM, Guiet Romain <[hidden email]> wrote: > Hi Pavica, > > I don’t have a direct solution to your problem, but it reminds me one > concern I had a few weeks ago (and that I overcame by a different way) and > thus I take advantage of your thread to make a suggestion of a new feature > to the CellCounter developper… > > Will be doable to have editable name for each counter ? > Using a text field (maybe limited in size). > If this configuration could be save and load (in text/csv file), it will > be even more usefull. > > Cheers, > > Romain > > > --------------------------------------------------------------- > Dr. Romain Guiet > Bioimaging and Optics Platform (PT-BIOP) > Ecole Polytechnique Fédérale de Lausanne (EPFL) > Faculty of Life Sciences > Station 19, AI 0140 > CH-1015 Lausanne > > Phone: [+4121 69] 39629 > http://biop.epfl.ch/ > --------------------------------------------------------------- > > ________________________________________ > De : ImageJ Interest Group [[hidden email]] de la part de Pavica [ > [hidden email]] > Envoyé : vendredi 27 février 2015 08:29 > À : [hidden email] > Objet : How to add more than 38 counter types in Cell Counter plugin > > Hi, > > I analyze invertebrate diversity on bottom-ocean images and I find many > invertebrate species on each photo. However, I can only add up to 38 > counter > types in the Cell Counter plugin since the number of the counter types that > fit on the screen depends on the size of my computer screen (!?) from the > top to the bottom of the screen. You can imagine that I am using the > biggest > screen I can find. > > I am wandering whether anyone ran into this problem and maybe solved it by > fixing the plugin so that you can scroll up and down the cell counter > window, once you ran into the bottom of the screen. > > I really probably need around 200 counter types, since I have several > stations and the species only partially overlap between the stations. So > far, I have been assigning Types 1-38 for different species at each station > but this eventually leads to a major confusion and a lot of additional work > on data editing. > > Thank you, > > Pavica > > > > -- > View this message in context: > http://imagej.1557.x6.nabble.com/How-to-add-more-than-38-counter-types-in-Cell-Counter-plugin-tp5011798.html > Sent from the ImageJ mailing list archive at Nabble.com. > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Thank you Curtis for the GitHub migration! It's always great to know that we can now work on the code more easily when needed!
Regarding this particular question would it not be easier to use the Multipoint tool along with the ROI manager? You can add as many multipoint ROIs as you want, rename them and so on. I made (really quickly) an actionBar that allows you to create as many point selections as you want, and lets you count the categories in the end. The categories can have whatever name you want. To make it work 1. Install Actiobar, a great tool by Jerome Mutterer. This allows us to make simple buttons for macros. http://imagejdocu.tudor.lu/doku.php?id=plugin:utilities:action_bar:start#downloads 2. Download the code from dropbox here https://www.dropbox.com/s/9p7elt8t6nn0hn8/Multi_Count.ijm?dl=0 3.Copy the ijm file into plugins/ActionBar folder Restard ImageJ/Fiji You will then find the macro under Plugins->ActionBar->Multi Count To use it 1. Open your image 2. Click on New category You will be prompted to create at least 1 point before clicking OK You will be asked for the name or it will be given a default 3. To add more points, Select the category in the ROI manager, Click to add as many points as you want, and finally click on Update Category 4. If AFTER you Updated there is a point you would like to remove, you can click on Delete Last Point 5. To Measure the counts in each category, hit Measure and you will have a table with each column that has the name you gave it You can rename categories anytime after the first step, their names are only used for the columns in when you are launching Measure. 6. To save the ROI set just hit Save Rois You can drag and drop the zip file into ImageJ to reopen the ROIs whenever you need to see them Right now we cannot pre-import a list of names but this is feasible if you think it could be interesting. Best Oli -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Hi Olivier,
> would it not be easier to use the Multipoint tool along with the ROI > manager? I would still strongly recommend automated or semi-automated counting if possible. But if that is truly impossible for some reason, then that is a great suggestion. > I made (really quickly) an actionBar that allows you to create as many > point selections as you want, and lets you count the categories in the > end. Thank you very much for that! > To make it work It would help users a lot to use an update site for distribution [1]. Distributing these things with Fiji is also a nice option [2]; just let me know if you want to pursue that. Regards, Curtis [1] http://imagej.net/Distribution [2] http://fiji.sc/Fiji_contribution_requirements On Tue, Mar 3, 2015 at 8:27 AM, Burri Olivier <[hidden email]> wrote: > Thank you Curtis for the GitHub migration! It's always great to know that > we can now work on the code more easily when needed! > > Regarding this particular question would it not be easier to use the > Multipoint tool along with the ROI manager? > > You can add as many multipoint ROIs as you want, rename them and so on. > > I made (really quickly) an actionBar that allows you to create as many > point selections as you want, and lets you count the categories in the end. > > The categories can have whatever name you want. > > To make it work > > 1. Install Actiobar, a great tool by Jerome Mutterer. This allows us to > make simple buttons for macros. > > http://imagejdocu.tudor.lu/doku.php?id=plugin:utilities:action_bar:start#downloads > > 2. Download the code from dropbox here > https://www.dropbox.com/s/9p7elt8t6nn0hn8/Multi_Count.ijm?dl=0 > > 3.Copy the ijm file into plugins/ActionBar folder > > Restard ImageJ/Fiji > > You will then find the macro under > Plugins->ActionBar->Multi Count > > To use it > 1. Open your image > 2. Click on New category > You will be prompted to create at least 1 point before clicking OK > You will be asked for the name or it will be given a default > > 3. To add more points, > Select the category in the ROI manager, > Click to add as many points as you want, and finally click on > Update Category > > 4. If AFTER you Updated there is a point you would like to remove, you can > click on > Delete Last Point > > 5. To Measure the counts in each category, hit > Measure and you will have a table with each column that has the > name you gave it > > > You can rename categories anytime after the first step, their names are > only used for the columns in when you are launching Measure. > > 6. To save the ROI set just hit > Save Rois > > > > You can drag and drop the zip file into ImageJ to reopen the ROIs whenever > you need to see them > > Right now we cannot pre-import a list of names but this is feasible if you > think it could be interesting. > > Best > > Oli > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Hi Curtis
> I would still strongly recommend automated or semi-automated counting if > possible. But if that is truly impossible for some reason, then that is a great > suggestion. I agree that some semi-automatic method could be very useful, however images like the ones processed by Pavica tend to have a lot of artifacts and uninteresting objects that are rather difficult to distinguish from useful ones. The only thing that I could think of is finding a way to rougly segment the images initially and then use something like Cell Profiler Analyst to hope and find some features to help classify the critters. The problem there is that without a dedicated image processing analyst the learning curve for a biologist to segment and is a bit steep until we can even assess whether the results are good or not... My two cents > > It would help users a lot to use an update site for distribution [1]. > Distributing these things with Fiji is also a nice option [2]; just let me know if > you want to pursue that. We do have an update site and we can now pack the actionBars into JAR files. I just did not think it would be so relevant in this case (I did write it in 30 minutes). But I could put it on GIT or in a GIST if you thing that would be useful. It is just macro code in the end, not a full plugin... > [1] http://imagej.net/Distribution > [2] http://fiji.sc/Fiji_contribution_requirements Thanks a lot for the references! Regarding update sites, our group has one now and we have a few plugins and actionbars on it. The problem I have now is that as we start producing more and more tools, having them all in the same update site makes it a bit bulky, as not all users want to get all our tools. Is there some suggestion as to how to address this? Should we split our packages in different update sites, for example? Thanks again and all the best Oli -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Hi Curtis and Oli,
thank you so much for helping with this. Oliver, thanks for explaining the problem of automated counting - I do have a lots of artifacts and things like rocks of different sizes, shapes and colors; mounds, pits, holes etc., or only a part of an animal sticking from the mud... I think all this would make the automatic counting of animals challenging. And besides all that, I am not a programmer, I am a biologist with a very very limited coding experience. That being said, I will try to implement your guys' suggestions by following your instructions, even though I've never done anything in ImageJ that involved more than merely pushing the buttons, so we'll see how it goes! Thank you! Pavica |
Hi Pavica,
> I've never done anything in ImageJ that involved more than merely > pushing the buttons I suggest giving the Trainable Weka Segmentation a try. It would only take a few minutes to try it out, and the worst-case scenario is that it can't handle your data. But it might save you loads of time. http://imagej.net/Trainable_Weka_Segmentation Regards, Curtis On Wed, Mar 4, 2015 at 5:21 PM, Pavica <[hidden email]> wrote: > Hi Curtis and Oli, > > thank you so much for helping with this. Oliver, thanks for explaining the > problem of automated counting - I do have a lots of artifacts and things > like rocks of different sizes, shapes and colors; mounds, pits, holes etc., > or only a part of an animal sticking from the mud... I think all this would > make the automatic counting of animals challenging. And besides all that, I > am not a programmer, I am a biologist with a very very limited coding > experience. That being said, I will try to implement your guys' suggestions > by following your instructions, even though I've never done anything in > ImageJ that involved more than merely pushing the buttons, so we'll see how > it goes! > > Thank you! > > Pavica > > > > -- > View this message in context: > http://imagej.1557.x6.nabble.com/How-to-add-more-than-38-counter-types-in-Cell-Counter-plugin-tp5011798p5011880.html > Sent from the ImageJ mailing list archive at Nabble.com. > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Free forum by Nabble | Edit this page |