Loading... |
Reply to author |
Edit post |
Move post |
Delete this post |
Delete this post and replies |
Change post date |
Print post |
Permalink |
Raw mail |
Hey there.
I'd like to find a way to get a segmentation mask automatically and it should be guided by two other inputs which I want to create in a preceding step. Please have a look at this picture: https://dl.dropbox.com/u/414433/imagej/auto-segmentation.jpg I managed to create a very robust outer mask of pixels which are not important (can be deleted/transparent). A second "inner mask" incorporates opaque pixels I want to keep. The wanted segmentation mask can be found between these two inside of the white area. There are several plugins for segmentation. Which one is best for my purposes? Any tips are appreciated. Best regards, Rainer -- Rainer M. Engel, Dipl. Digital Artist scientific|Media GbR Pichelsdorfer Str. 143 13595 Berlin -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Loading... |
Reply to author |
Edit post |
Move post |
Delete this post |
Delete this post and replies |
Change post date |
Print post |
Permalink |
Raw mail |
Hi Rainer,
> There are several plugins for segmentation. Which one is best for my > purposes? It really depends on your data. The simplest segmentation technique is thresholding (Image > Adjust > Threshold) with one of the built-in auto-thresholding methods. Steps would be: 1) Generate your inner and outer masks. 2) AND the inner mask to your image (using the Image Calculator). 3) AND the outer mask to your image. You should now have an image that is all zeroes except for the area between the masks. (If not, invert one or both masks first before ANDing.) 4) Run the Auto Threshold plugin on the resultant image with the desired method, checking the "Ignore black" or "Ignore white" and "White objects on black background" boxes as appropriate. (You can also choose the "Try all" method to get a nice big side-by-side comparison of results.) 5) Recombine the resultant mask (produced by Auto Threshold) with your inner mask, outer mask and/or original data as desired for further processing. If you have very noisy data, you could alternately try the Advanced Weka Segmentation Editor to use machine learning on the ambiguous area, rather than simple thresholding. If you go that route, you can mark everything in your inner mask as class 1, and everything in your outer mask as class 2, which should give it plenty of fodder for further deduction. The complete list of segmentation plugins available in Fiji can be browsed here: http://fiji.sc/Category:Segmentation Regards, Curtis On Thu, Feb 21, 2013 at 11:24 AM, Rainer M. Engel <[hidden email]> wrote: > Hey there. > > I'd like to find a way to get a segmentation mask automatically and it > should be guided by two other inputs which I want to create in a > preceding step. > > Please have a look at this picture: > https://dl.dropbox.com/u/414433/imagej/auto-segmentation.jpg > > I managed to create a very robust outer mask of pixels which are not > important (can be deleted/transparent). A second "inner mask" > incorporates opaque pixels I want to keep. The wanted segmentation mask > can be found between these two inside of the white area. > > There are several plugins for segmentation. Which one is best for my > purposes? Any tips are appreciated. > > Best regards, > Rainer > > > -- > Rainer M. Engel, Dipl. Digital Artist > scientific|Media GbR > Pichelsdorfer Str. 143 > 13595 Berlin > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > ... [show rest of quote] -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Loading... |
Reply to author |
Edit post |
Move post |
Delete this post |
Delete this post and replies |
Change post date |
Print post |
Permalink |
Raw mail |
Dear Curtis,
thank you for your detailed information. I'll definitely follow this description. Best Regards, Rainer Am 21.02.2013 21:17, schrieb Curtis Rueden: > Hi Rainer, > >> There are several plugins for segmentation. Which one is best for my >> purposes? > > It really depends on your data. The simplest segmentation technique is > thresholding (Image > Adjust > Threshold) with one of the built-in > auto-thresholding methods. Steps would be: > > 1) Generate your inner and outer masks. > 2) AND the inner mask to your image (using the Image Calculator). > 3) AND the outer mask to your image. > > You should now have an image that is all zeroes except for the area between > the masks. (If not, invert one or both masks first before ANDing.) > > 4) Run the Auto Threshold plugin on the resultant image with the desired > method, checking the "Ignore black" or "Ignore white" and "White objects on > black background" boxes as appropriate. (You can also choose the "Try all" > method to get a nice big side-by-side comparison of results.) > > 5) Recombine the resultant mask (produced by Auto Threshold) with your > inner mask, outer mask and/or original data as desired for further > processing. > > If you have very noisy data, you could alternately try the Advanced Weka > Segmentation Editor to use machine learning on the ambiguous area, rather > than simple thresholding. If you go that route, you can mark everything in > your inner mask as class 1, and everything in your outer mask as class 2, > which should give it plenty of fodder for further deduction. > > The complete list of segmentation plugins available in Fiji can be browsed > here: > http://fiji.sc/Category:Segmentation > > Regards, > Curtis > > > On Thu, Feb 21, 2013 at 11:24 AM, Rainer M. Engel <[hidden email]> wrote: > >> Hey there. >> >> I'd like to find a way to get a segmentation mask automatically and it >> should be guided by two other inputs which I want to create in a >> preceding step. >> >> Please have a look at this picture: >> https://dl.dropbox.com/u/414433/imagej/auto-segmentation.jpg >> >> I managed to create a very robust outer mask of pixels which are not >> important (can be deleted/transparent). A second "inner mask" >> incorporates opaque pixels I want to keep. The wanted segmentation mask >> can be found between these two inside of the white area. >> >> There are several plugins for segmentation. Which one is best for my >> purposes? Any tips are appreciated. >> >> Best regards, >> Rainer >> >> >> -- >> Rainer M. Engel, Dipl. Digital Artist >> scientific|Media GbR >> Pichelsdorfer Str. 143 >> 13595 Berlin >> >> -- >> ImageJ mailing list: http://imagej.nih.gov/ij/list.html >> > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > ... [show rest of quote] -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Free forum by Nabble | Disable Popup Ads | Edit this page |