SIFT features and stats
Locked
 
|
Hi imagej gurus,
I'm rather new to imagej and fiji however it has already been proven to be a very useful platform for me as such I would like to continue using fiji / imagej in a proof of concept registration exercise. But, I need some advice on getting to certain data. About the image vision problem: I want to align webcam images and compensate for image shifts by registering images in a sliding window -through a stack- approach. The reason for doing this with a sliding window approach is that features are not stable due to changes in vegetation state (leaf on / leaf off). Within this context I would like to extract feature statistics from the 'Extract SIFT Correspondences' plugin. Sadly, although I can read the features in to bUnwarpJ I cannot access the statistics as produced by the SIFT procedure as there is no text output. I need access to these statistics as they are a marker for bad images (snow on the lens, mist, other reasons why features could not be detected). Using these statistics I can kick out, 'bad' days and only register the 'good' days. One option would be to scrobble the output to the terminal output for any log data, but this seems cumbersome and non robust. Is there another way to use standard fiji macro language and access this log data? And if not are these statistics on SIFT analysis readable in some other setup (jython / java calling of the classes). Any help or input on how to tackle this is appreciated. Cheers, Koen -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
|
|
Am 04.11.2013 21:25, schrieb Koen Hufkens:
> Sadly, although I can read the > features in to bUnwarpJ I cannot access the statistics as produced by the > SIFT procedure as there is no text output. Well you could write one yourself with a script language, or the built in macro language since SIFT creates a multi point selection you could read the position of these points. //+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ //run SIFT between two images (also poss. from stack) type = selectionType(); if (type==10) { getSelectionCoordinates(xArr, yArr); print("SIFT found: "+xArr.length+" corresponding points"); print("printing coordinates:"); for (i=0; i<xArr.length; i++) { showProgress(i+1, xArr.length); print(xArr[i]+","+yArr[i]); } } else { print("No correspondencies could be found."); } //+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 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 |
|
|
Thanks Rainer,
I'm new to ImageJ so I'm really unfamiliar with the options and syntax / available functions. Cheers, Koen On Tue, Nov 5, 2013 at 4:46 AM, Rainer M. Engel <[hidden email]> wrote: > Am 04.11.2013 21:25, schrieb Koen Hufkens: > > Sadly, although I can read the > > features in to bUnwarpJ I cannot access the statistics as produced by the > > SIFT procedure as there is no text output. > > Well you could write one yourself with a script language, or the built > in macro language since SIFT creates a multi point selection you could > read the position of these points. > > //+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > //run SIFT between two images (also poss. from stack) > > type = selectionType(); > if (type==10) { > getSelectionCoordinates(xArr, yArr); > print("SIFT found: "+xArr.length+" corresponding points"); > print("printing coordinates:"); > for (i=0; i<xArr.length; i++) { > showProgress(i+1, xArr.length); > print(xArr[i]+","+yArr[i]); > } > } else { > print("No correspondencies could be found."); > } > //+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > 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 |
|
|
> I'm new to ImageJ
I'm sure everyone once was .. great you came around. The macro language of ImageJ is rather simple and powerful at the same time. I'm sure you will solve your problem. //otherwise if (questionsArise == true) { comeBack = "any time"; } Regards, Rainer Am 05.11.2013 17:13, schrieb Koen Hufkens: > Thanks Rainer, > > I'm new to ImageJ so I'm really unfamiliar with the options and syntax / > available functions. > > Cheers, > Koen > > > > > On Tue, Nov 5, 2013 at 4:46 AM, Rainer M. Engel <[hidden email]> wrote: > >> Am 04.11.2013 21:25, schrieb Koen Hufkens: >>> Sadly, although I can read the >>> features in to bUnwarpJ I cannot access the statistics as produced by the >>> SIFT procedure as there is no text output. >> >> Well you could write one yourself with a script language, or the built >> in macro language since SIFT creates a multi point selection you could >> read the position of these points. >> >> //+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >> //run SIFT between two images (also poss. from stack) >> >> type = selectionType(); >> if (type==10) { >> getSelectionCoordinates(xArr, yArr); >> print("SIFT found: "+xArr.length+" corresponding points"); >> print("printing coordinates:"); >> for (i=0; i<xArr.length; i++) { >> showProgress(i+1, xArr.length); >> print(xArr[i]+","+yArr[i]); >> } >> } else { >> print("No correspondencies could be found."); >> } >> //+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >> >> 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 > > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
«
Return to ImageJ
|
1 view|%1 views
| Free forum by Nabble | Edit this page |