Measure Motion / Blob Detection
Hello all, I had two questions: 1) I see there is now the option in measure area to ignore dusted objects, is it possible to have this feature in the measure motion as well? 2) For measure and mark blobs, Is there any way to use the output of the measure motion function to link the blobs since the triangle number changes in each frame, but following measure motion we know where it goes in the next frame. Thanks, Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu> [South Dakota Mines]<https://www.sdsmt.edu/> [South Dakota Mines on Facebook]<https://www.facebook.com/SouthDakotaMines/>[South Dakota Mines on Instagram]<https://www.instagram.com/southdakotamines/>[South Dakota Mines on Twitter]<https://twitter.com/sdsmt>[South Dakota Mines on Snapchat]<https://www.snapchat.com/add/sdsmt>
Hi Brandon, Yes, the "measure area" command has the includeMasked option default false so the area of the small blobs masked by the surface dust command is not included in the calculation https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#area <https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#area> The "measure motion" command shows line segments (hairs, prickles) on a surface showing which direction a surface is moving in a series of surfaces, for instance in a surface depiction of a 3D light microscopy time series. Are you saying that measure motion shows the line segments even for blobs hidden with "surface dust", and that you want an option to not show the line segments for those hidden blobs? The measure motion command is a rarely used feature and probably no one else ever tried it with hiding dust https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#motion <https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#motion> It is possible to make that improvement but hard to find time to do it. For tracking the surface as it moves, which points at time t=2 correspond to which points at t=1, that is a complex problem. Maybe the measure motion vectors are a tiny part of a solution. This would produce a large quantity of data matching many of the surface vertices between every two time points. An critical consideration would be how you are planning to use that information. In general tracking surfaces that split, merge, disappear, ... and analyzing requires a lot of custom written software. We don't have the funding to supply special purpose surface tracking software so I think it is likely you would need to find other software that can do the job, probably that does not exist, so you would need a programmer to develop it for you. Maybe what you want it simpler than I think. But you would need to explain it in detail. Tom
On Oct 1, 2020, at 10:20 AM, Scott, Brandon L. <Brandon.Scott@sdsmt.edu> wrote:
Hello all, I had two questions:
1) I see there is now the option in measure area to ignore dusted objects, is it possible to have this feature in the measure motion as well? 2) For measure and mark blobs, Is there any way to use the output of the measure motion function to link the blobs since the triangle number changes in each frame, but following measure motion we know where it goes in the next frame.
Thanks, Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu <mailto:Brandon.Scott@sdsmt.edu> <https://www.sdsmt.edu/> <https://www.facebook.com/SouthDakotaMines/> <https://www.instagram.com/southdakotamines/> <https://twitter.com/sdsmt> <https://www.snapchat.com/add/sdsmt> _______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu <mailto:ChimeraX-users@cgl.ucsf.edu> Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users <https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users>
Hi Tom, Thanks for the response. I was not very clear in what I was asking, my apologies. I am trying to record the centroids of any displayed connected components. I would then be able to use trackpy for the actual tracking. Markers-> connected_center does a great job of quickly doing this, but I am wondering if there is a way do it for all the connected components without having to manually specify a point like how the Blob_Masker class calculates the area. Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu> [South Dakota Mines]<https://www.sdsmt.edu/> [South Dakota Mines on Facebook]<https://www.facebook.com/SouthDakotaMines/>[South Dakota Mines on Instagram]<https://www.instagram.com/southdakotamines/>[South Dakota Mines on Twitter]<https://twitter.com/sdsmt>[South Dakota Mines on Snapchat]<https://www.snapchat.com/add/sdsmt> ________________________________ From: Tom Goddard <goddard@sonic.net> Sent: Thursday, October 1, 2020 3:59 PM To: Scott, Brandon L. <Brandon.Scott@sdsmt.edu> Cc: chimerax-users@cgl.ucsf.edu <chimerax-users@cgl.ucsf.edu> Subject: [EXT] Re: [chimerax-users] Measure Motion / Blob Detection *** This email is from an EXTERNAL sender. Use CAUTION before opening attachments or clicking links.*** Hi Brandon, Yes, the "measure area" command has the includeMasked option default false so the area of the small blobs masked by the surface dust command is not included in the calculation https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#area The "measure motion" command shows line segments (hairs, prickles) on a surface showing which direction a surface is moving in a series of surfaces, for instance in a surface depiction of a 3D light microscopy time series. Are you saying that measure motion shows the line segments even for blobs hidden with "surface dust", and that you want an option to not show the line segments for those hidden blobs? The measure motion command is a rarely used feature and probably no one else ever tried it with hiding dust https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#motion It is possible to make that improvement but hard to find time to do it. For tracking the surface as it moves, which points at time t=2 correspond to which points at t=1, that is a complex problem. Maybe the measure motion vectors are a tiny part of a solution. This would produce a large quantity of data matching many of the surface vertices between every two time points. An critical consideration would be how you are planning to use that information. In general tracking surfaces that split, merge, disappear, ... and analyzing requires a lot of custom written software. We don't have the funding to supply special purpose surface tracking software so I think it is likely you would need to find other software that can do the job, probably that does not exist, so you would need a programmer to develop it for you. Maybe what you want it simpler than I think. But you would need to explain it in detail. Tom On Oct 1, 2020, at 10:20 AM, Scott, Brandon L. <Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu>> wrote: Hello all, I had two questions: 1) I see there is now the option in measure area to ignore dusted objects, is it possible to have this feature in the measure motion as well? 2) For measure and mark blobs, Is there any way to use the output of the measure motion function to link the blobs since the triangle number changes in each frame, but following measure motion we know where it goes in the next frame. Thanks, Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu> [South Dakota Mines]<https://www.sdsmt.edu/> [South Dakota Mines on Facebook]<https://www.facebook.com/SouthDakotaMines/>[South Dakota Mines on Instagram]<https://www.instagram.com/southdakotamines/>[South Dakota Mines on Twitter]<https://twitter.com/sdsmt>[South Dakota Mines on Snapchat]<https://www.snapchat.com/add/sdsmt> _______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu<mailto:ChimeraX-users@cgl.ucsf.edu> Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users
Hi Brandon, Ok, I added a "mark connected" command to ChimeraX that places a spherical marker on every connected surface piece. If you are hiding dust it only uses the displayed surface pieces. Here's an example open 1a0m from eds mark connected #1 radius 0.2 color gold There is a mouse mode to place a marker on a clicked surface piece under the Markers toolbar, and there is the surface dust command that finds all connected pieces and hides the small ones. I just looked over the Python code for those and combined them to add this new command. It is a good example of the power of using Python to call existing ChimeraX capabilities. So I made a ChimeraX recipe web page showing how it was done. https://rbvi.github.io/chimerax-recipes/mark_blobs/mark_blobs.html Tom
On Oct 9, 2020, at 12:18 PM, Scott, Brandon L. <Brandon.Scott@sdsmt.edu> wrote:
Hi Tom,
Thanks for the response. I was not very clear in what I was asking, my apologies.
I am trying to record the centroids of any displayed connected components. I would then be able to use trackpy for the actual tracking. Markers-> connected_center does a great job of quickly doing this, but I am wondering if there is a way do it for all the connected components without having to manually specify a point like how the Blob_Masker class calculates the area.
Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu <mailto:Brandon.Scott@sdsmt.edu> <https://www.sdsmt.edu/> <https://www.facebook.com/SouthDakotaMines/> <https://www.instagram.com/southdakotamines/> <https://twitter.com/sdsmt> <https://www.snapchat.com/add/sdsmt> From: Tom Goddard <goddard@sonic.net <mailto:goddard@sonic.net>> Sent: Thursday, October 1, 2020 3:59 PM To: Scott, Brandon L. <Brandon.Scott@sdsmt.edu <mailto:Brandon.Scott@sdsmt.edu>> Cc: chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu> <chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu>> Subject: [EXT] Re: [chimerax-users] Measure Motion / Blob Detection
*** This email is from an EXTERNAL sender. Use CAUTION before opening attachments or clicking links.*** Hi Brandon,
Yes, the "measure area" command has the includeMasked option default false so the area of the small blobs masked by the surface dust command is not included in the calculation
https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#area <https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#area>
The "measure motion" command shows line segments (hairs, prickles) on a surface showing which direction a surface is moving in a series of surfaces, for instance in a surface depiction of a 3D light microscopy time series. Are you saying that measure motion shows the line segments even for blobs hidden with "surface dust", and that you want an option to not show the line segments for those hidden blobs? The measure motion command is a rarely used feature and probably no one else ever tried it with hiding dust
https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#motion <https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#motion>
It is possible to make that improvement but hard to find time to do it.
For tracking the surface as it moves, which points at time t=2 correspond to which points at t=1, that is a complex problem. Maybe the measure motion vectors are a tiny part of a solution. This would produce a large quantity of data matching many of the surface vertices between every two time points. An critical consideration would be how you are planning to use that information. In general tracking surfaces that split, merge, disappear, ... and analyzing requires a lot of custom written software.
We don't have the funding to supply special purpose surface tracking software so I think it is likely you would need to find other software that can do the job, probably that does not exist, so you would need a programmer to develop it for you. Maybe what you want it simpler than I think. But you would need to explain it in detail.
Tom
On Oct 1, 2020, at 10:20 AM, Scott, Brandon L. <Brandon.Scott@sdsmt.edu <mailto:Brandon.Scott@sdsmt.edu>> wrote:
Hello all, I had two questions:
1) I see there is now the option in measure area to ignore dusted objects, is it possible to have this feature in the measure motion as well? 2) For measure and mark blobs, Is there any way to use the output of the measure motion function to link the blobs since the triangle number changes in each frame, but following measure motion we know where it goes in the next frame.
Thanks, Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu <mailto:Brandon.Scott@sdsmt.edu> <https://www.sdsmt.edu/> <https://www.facebook.com/SouthDakotaMines/> <https://www.instagram.com/southdakotamines/> <https://twitter.com/sdsmt> <https://www.snapchat.com/add/sdsmt> _______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu <mailto:ChimeraX-users@cgl.ucsf.edu> Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users <https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users>
ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu <mailto:ChimeraX-users@cgl.ucsf.edu> Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users <https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users>
Thank you! That is exactly what I was trying to figure out. I really appreciate you help and the recipe you put up, that makes things a bit clearer on how to move forward with some other ideas. -Brandon ________________________________ From: Tom Goddard <goddard@sonic.net> Sent: Friday, October 9, 2020 9:27 PM To: Scott, Brandon L. <Brandon.Scott@sdsmt.edu> Cc: chimerax-users@cgl.ucsf.edu <chimerax-users@cgl.ucsf.edu> Subject: Re: [chimerax-users] [EXT] Re: Measure Motion / Blob Detection *** This email is from an EXTERNAL sender. Use CAUTION before opening attachments or clicking links.*** Hi Brandon, Ok, I added a "mark connected" command to ChimeraX that places a spherical marker on every connected surface piece. If you are hiding dust it only uses the displayed surface pieces. Here's an example open 1a0m from eds mark connected #1 radius 0.2 color gold There is a mouse mode to place a marker on a clicked surface piece under the Markers toolbar, and there is the surface dust command that finds all connected pieces and hides the small ones. I just looked over the Python code for those and combined them to add this new command. It is a good example of the power of using Python to call existing ChimeraX capabilities. So I made a ChimeraX recipe web page showing how it was done. https://rbvi.github.io/chimerax-recipes/mark_blobs/mark_blobs.html Tom [cid:A4BC37C8-4C49-48EE-A2D0-781F3B87DFEC] On Oct 9, 2020, at 12:18 PM, Scott, Brandon L. <Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu>> wrote: Hi Tom, Thanks for the response. I was not very clear in what I was asking, my apologies. I am trying to record the centroids of any displayed connected components. I would then be able to use trackpy for the actual tracking. Markers-> connected_center does a great job of quickly doing this, but I am wondering if there is a way do it for all the connected components without having to manually specify a point like how the Blob_Masker class calculates the area. Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu> [South Dakota Mines]<https://www.sdsmt.edu/> [South Dakota Mines on Facebook]<https://www.facebook.com/SouthDakotaMines/>[South Dakota Mines on Instagram]<https://www.instagram.com/southdakotamines/>[South Dakota Mines on Twitter]<https://twitter.com/sdsmt>[South Dakota Mines on Snapchat]<https://www.snapchat.com/add/sdsmt> ________________________________ From: Tom Goddard <goddard@sonic.net<mailto:goddard@sonic.net>> Sent: Thursday, October 1, 2020 3:59 PM To: Scott, Brandon L. <Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu>> Cc: chimerax-users@cgl.ucsf.edu<mailto:chimerax-users@cgl.ucsf.edu> <chimerax-users@cgl.ucsf.edu<mailto:chimerax-users@cgl.ucsf.edu>> Subject: [EXT] Re: [chimerax-users] Measure Motion / Blob Detection *** This email is from an EXTERNAL sender. Use CAUTION before opening attachments or clicking links.*** Hi Brandon, Yes, the "measure area" command has the includeMasked option default false so the area of the small blobs masked by the surface dust command is not included in the calculation https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#area The "measure motion" command shows line segments (hairs, prickles) on a surface showing which direction a surface is moving in a series of surfaces, for instance in a surface depiction of a 3D light microscopy time series. Are you saying that measure motion shows the line segments even for blobs hidden with "surface dust", and that you want an option to not show the line segments for those hidden blobs? The measure motion command is a rarely used feature and probably no one else ever tried it with hiding dust https://www.cgl.ucsf.edu/chimerax/docs/user/commands/measure.html#motion It is possible to make that improvement but hard to find time to do it. For tracking the surface as it moves, which points at time t=2 correspond to which points at t=1, that is a complex problem. Maybe the measure motion vectors are a tiny part of a solution. This would produce a large quantity of data matching many of the surface vertices between every two time points. An critical consideration would be how you are planning to use that information. In general tracking surfaces that split, merge, disappear, ... and analyzing requires a lot of custom written software. We don't have the funding to supply special purpose surface tracking software so I think it is likely you would need to find other software that can do the job, probably that does not exist, so you would need a programmer to develop it for you. Maybe what you want it simpler than I think. But you would need to explain it in detail. Tom On Oct 1, 2020, at 10:20 AM, Scott, Brandon L. <Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu>> wrote: Hello all, I had two questions: 1) I see there is now the option in measure area to ignore dusted objects, is it possible to have this feature in the measure motion as well? 2) For measure and mark blobs, Is there any way to use the output of the measure motion function to link the blobs since the triangle number changes in each frame, but following measure motion we know where it goes in the next frame. Thanks, Brandon Scott, PhD CZI Imaging Scientist, Research Assistant Professor Nanoscience & Nanoengineering South Dakota Mines 501 E. Saint Joseph St., Rapid City, SD 57701 724.510.1253 | Brandon.Scott@sdsmt.edu<mailto:Brandon.Scott@sdsmt.edu> [South Dakota Mines]<https://www.sdsmt.edu/> [South Dakota Mines on Facebook]<https://www.facebook.com/SouthDakotaMines/>[South Dakota Mines on Instagram]<https://www.instagram.com/southdakotamines/>[South Dakota Mines on Twitter]<https://twitter.com/sdsmt>[South Dakota Mines on Snapchat]<https://www.snapchat.com/add/sdsmt> _______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu<mailto:ChimeraX-users@cgl.ucsf.edu> Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users _______________________________________________ ChimeraX-users mailing list ChimeraX-users@cgl.ucsf.edu<mailto:ChimeraX-users@cgl.ucsf.edu> Manage subscription: https://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users
participants (2)
-
Scott, Brandon L. -
Tom Goddard