Area measurement
Locked
 
|
Dear experts
I have a big image sequence coming from micro-tomografy analysis. I need to calculate the area of some structures that have a gery value very similar to other structures (see the picture, I need the thin layer between arrows). So when I spotlight this structures by Histogram and Thereshold, I can not separate it from other parts that have similar grey values. Finally the measurament is terribly overstimate. Do you know some trick to resolve my problem? Thank you Francesco mct_fra.jpg |
|
|
Francesco,
for a first try I've applied two times "Process>Noise>Despeckle" to your JPG-sample. Then I've chosen a threshold at value 112. With this setting a separation is possible but of course you will have to set the correct threshold for every image by hand. Furthermore, you will have to close the ends of your "bands of interest" by hand and then apply the wand tool. >Dear experts > >I have a big image sequence coming from micro-tomografy analysis. I need to >calculate the area of some structures that have a gery value very similar to >other structures (see the picture, I need the thin layer between arrows). So >when I spotlight this structures by Histogram and Thereshold, I can not >separate it from other parts that have similar grey values. Finally the >measurament is terribly overstimate. >Do you know some trick to resolve my problem? > >Thank you > >Francesco >http://n2.nabble.com/file/n3598079/mct_fra.jpg mct_fra.jpg >-- >View this message in context: >http://n2.nabble.com/Area-measurement-tp3598079p3598079.html >Sent from the ImageJ mailing list archive at Nabble.com. HTH -- Herbie ------------------------ <http://www.gluender.de> |
|
|
Thank you for your quick response!
unfortunately I still have the problem beacause a lot of structures in the full picture have values around 112 and furthermore I have more than 1300 slices for sample....and every slice have little different value for the "bands". You can find for example 99, 103, 112........ Francesco
|
|
|
Francesco,
I fear you won't find an approach for the automatic isolation of the "bands". You'll always need to manually adjust some parameters. >Thank you for your quick response! > >unfortunately I still have the problem beacause a lot of structures in the >full picture have values around 112 and furthermore I have more than 1300 >slices for sample....and every slice have little different value for the >"bands". You can find for example 99, 103, 112........ I wrote: "... of course you will have to set the correct threshold for every image by hand." >Francesco >Gluender-3 wrote: >> >> Francesco, >> >> for a first try I've applied two times "Process>Noise>Despeckle" to >> your JPG-sample. Then I've chosen a threshold at value 112. With this > > setting a separation is possible but of course you will have to set > > the correct threshold for every image by hand. >> >> Furthermore, you will have to close the ends of your "bands of >> interest" by hand and then apply the wand tool. >> >>>Dear experts >>> >>>I have a big image sequence coming from micro-tomografy analysis. I need >to >>>calculate the area of some structures that have a gery value very similar >to >>>other structures (see the picture, I need the thin layer between arrows). >So >>>when I spotlight this structures by Histogram and Thereshold, I can not >>>separate it from other parts that have similar grey values. Finally the >>>measurament is terribly overstimate. >>>Do you know some trick to resolve my problem? >>> >>>Thank you >>> >>>Francesco > >>http://n2.nabble.com/file/n3598079/mct_fra.jpg mct_fra.jpg > > > HTH >> -- >> >> Herbie >> >> ------------------------ > > <http://www.gluender.de> > In fact there is more than human vision that you would need to solve the problem, you'll need specific human knowledge as well... For instance to solve the problem: Where do the bands end? Best -- Herbie ------------------------ <http://www.gluender.de> |
|
|
If I will find a method to do this automatically I will be crazy for happiness!
But I will be happy even finding some resolution for only one slice. Anyway, even filtering, or using Despeckle, the problem still exist. I think that I have to find some command or plugin to differentiate to distinct material that have similar grey value Regard your question, the "bands" are covering of gypsum around calcite grains. And every grain have different thickness of gypsum. Every picture represent a slice perpendicular to z axis of a reactor.I sent a detail because in normal size you can not see the bands Best regards Francesco
|
|
|
Hi Francesco,
to me it looks like a problem where image processing won't help, I would rather suggest a solution based on an improved method of image recording. If this is microtomography using x rays with rather low energy, maybe you could have two different filters and subtract the results, to get some contrast from the sulfur K edge (2472 eV)? This should make it more easy to discriminate between gypsum and calcite. Michael ________________________________________________________________ On 8 Sep 2009, at 10:46, fra-off wrote: > If I will find a method to do this automatically I will be crazy for > happiness! > But I will be happy even finding some resolution for only one slice. > Anyway, even filtering, or using Despeckle, the problem still exist. > I think that I have to find some command or plugin to differentiate to > distinct material that have similar grey value > Regard your question, the "bands" are covering of gypsum around > calcite > grains. And every grain have different thickness of gypsum. Every > picture > represent a slice perpendicular to z axis of a reactor.I sent a detail > because in normal size you can not see the bands > > Best regards > > Francesco |
|
|
Thanks
I'm tring to reconstruct the origninal file by octopus will see Francesco
|
«
Return to ImageJ
|
1 view|%1 views
| Free forum by Nabble | Edit this page |