46
General / Re: Orthomosaic generation - Mesh vs DEM
« on: September 12, 2017, 04:50:25 PM »
Thanks JMR - a very good explanation.
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
47
General / Orthomosaic generation - Mesh vs DEM
« on: September 11, 2017, 11:25:04 AM »
I'm now generating orthomosaics from every model I produce. The ability to review an underwater site in the comfort of your own home (and not get wet) is a great way of investigating what is on the seabed.
Example:
http://deep3d.co.uk/2017/08/23/bolt-bay-p47-d-crash-site/
When generating an orthomosaic there are two choices as to what to base the results on:
Mesh
DEM
I'm running a very large orthomosaic now (5 acres of shipwreck - 1 pix = 7.5mm) otherwise I would test it, but what are the differences? Any advantages over using the DEM vs the Mesh?
I am guessing it would be possible to generate a DEM from the sparse cloud and then derive the orthophoto from that, missing out the dense cloud and mesh, but would be interested to hear the experiences of others?
Example:
http://deep3d.co.uk/2017/08/23/bolt-bay-p47-d-crash-site/
When generating an orthomosaic there are two choices as to what to base the results on:
Mesh
DEM
I'm running a very large orthomosaic now (5 acres of shipwreck - 1 pix = 7.5mm) otherwise I would test it, but what are the differences? Any advantages over using the DEM vs the Mesh?
I am guessing it would be possible to generate a DEM from the sparse cloud and then derive the orthophoto from that, missing out the dense cloud and mesh, but would be interested to hear the experiences of others?
48
Python and Java API / Re: Split in Chunks - Copy Model
« on: September 03, 2017, 01:20:20 PM »
Thanks Alexey - much appreciated.
49
Python and Java API / Split in Chunks - Copy Model
« on: August 26, 2017, 03:46:49 PM »
Scripting newbie question:
I use the split_in_chunks script quite a lot (thank you to whoever authored it!) and its very useful.
I now have a situation where I want to split a chunk into separate chunks, duplicating not only the dense cloud/cameras but the 3D model as well.
I *think* I have found the section in the script:
new_chunk = chunk.copy()
new_chunk.label = "Chunk "+ str(i)+ "\\" + str(j) + "\\" + str(k)
new_chunk.model = None
doc.addChunk(new_chunk)
With the command that deals with the model in bold. I have checked the reference manual, but cannot find an example? I am guessing the None needs to be something else, and have tried a few guesses but no luck.
Can anyone help?
I use the split_in_chunks script quite a lot (thank you to whoever authored it!) and its very useful.
I now have a situation where I want to split a chunk into separate chunks, duplicating not only the dense cloud/cameras but the 3D model as well.
I *think* I have found the section in the script:
new_chunk = chunk.copy()
new_chunk.label = "Chunk "+ str(i)+ "\\" + str(j) + "\\" + str(k)
new_chunk.model = None
doc.addChunk(new_chunk)
With the command that deals with the model in bold. I have checked the reference manual, but cannot find an example? I am guessing the None needs to be something else, and have tried a few guesses but no luck.
Can anyone help?
50
Python and Java API / Re: Newbie - error running a script.
« on: April 28, 2017, 02:45:49 PM »
Hi Alexey
Script file is attached/uploaded.
Script file is attached/uploaded.
51
Python and Java API / Re: Newbie - error running a script.
« on: April 23, 2017, 12:01:21 PM »
Hi Alexey - here's the output:-
Out[1]: 2017-04-23 10:00:35 PySide2.QtWidgets.QDial
Out[1]: 2017-04-23 10:00:35 PySide2.QtWidgets.QDial
52
Python and Java API / Re: Newbie - error running a script.
« on: April 22, 2017, 09:32:43 PM »
Hi Alexey - yes the entire script including the line you mention.
The error occurs very soon on loading, triggered on the very line you refer to I believe.
The error occurs very soon on loading, triggered on the very line you refer to I believe.
53
Python and Java API / Newbie - error running a script.
« on: April 13, 2017, 08:01:55 PM »
Having recently upgraded to 1.3.1.4 I now find the following script:-
http://wiki.agisoft.com/w/index.php?title=Split_in_chunks.py&oldid=207
This script is really, really useful. Its a godsend for larger models such as this one:-
https://sketchfab.com/models/99d9f126ea0647f79e761d3cd2c1fca0
Created over 6 or 7 dives and now has nearly 10k images.
However, running the script triggers the following errors in the console:-
2017-04-13 17:47:31 class SplitDlg(QtWidgets.QDialog):
2017-04-13 17:47:31 NameError: name 'QtWidgets' is not defined
Can anyone suggest a fix for this one? its making me tear my hair out!
http://wiki.agisoft.com/w/index.php?title=Split_in_chunks.py&oldid=207
This script is really, really useful. Its a godsend for larger models such as this one:-
https://sketchfab.com/models/99d9f126ea0647f79e761d3cd2c1fca0
Created over 6 or 7 dives and now has nearly 10k images.
However, running the script triggers the following errors in the console:-
2017-04-13 17:47:31 class SplitDlg(QtWidgets.QDialog):
2017-04-13 17:47:31 NameError: name 'QtWidgets' is not defined
Can anyone suggest a fix for this one? its making me tear my hair out!
54
General / Re: Nikon D600 or D700
« on: May 30, 2016, 10:55:06 AM »
I use a housed D700 underwater. This has two issues; low light and particulates in the water that when lit with flash cause an effect called backscatter.
The D700 is not a current model, but has excellent high ISO characteristics and I regularly use ISO 1600 or 2000 to scan a site. The main benefit of high ISO is less flash power...which in turn means less backscatter.
And high ISO means you can shoot with torchlight instead of a flash gun.
An example lit with an Orcalight Seawolf torch:-
https://sketchfab.com/models/3e066c87ca8c44a69ff32b9e55f8d25b
An example lit with underwater flash guns:-
https://sketchfab.com/models/ed5190b32d514bc199cc00d85d4a9b19
An example of mixing both. The Abbot self propelled gun and Stalwart are lit with flash. The aircraft and Saracen lit with the Orcalight Seawolf. Four separate dives weeks apart, two divers and two different cameras were combined in this model.
https://sketchfab.com/models/c8743258e47747d8b319391cbbec81ec
So I would say the D700 is fine and has one significant advantage; its not a current camera and in spite of being very capable tool can be bought for a fraction of what the current cameras cost.
The D700 is not a current model, but has excellent high ISO characteristics and I regularly use ISO 1600 or 2000 to scan a site. The main benefit of high ISO is less flash power...which in turn means less backscatter.
And high ISO means you can shoot with torchlight instead of a flash gun.
An example lit with an Orcalight Seawolf torch:-
https://sketchfab.com/models/3e066c87ca8c44a69ff32b9e55f8d25b
An example lit with underwater flash guns:-
https://sketchfab.com/models/ed5190b32d514bc199cc00d85d4a9b19
An example of mixing both. The Abbot self propelled gun and Stalwart are lit with flash. The aircraft and Saracen lit with the Orcalight Seawolf. Four separate dives weeks apart, two divers and two different cameras were combined in this model.
https://sketchfab.com/models/c8743258e47747d8b319391cbbec81ec
So I would say the D700 is fine and has one significant advantage; its not a current camera and in spite of being very capable tool can be bought for a fraction of what the current cameras cost.
55
General / Hosting and Publishing Orthophotos
« on: April 24, 2016, 06:24:31 PM »
So I have generated a few orthophotos of some underwater 3D models.
The results are very impressive, and worthy of publication, but both Mpabox and 4DMapper (the two API-ready hosts that come with Photoscan Pro) are really geared around GPS/mapping.
Mixing underwater and GPS is possible (and we are working on it) but in absence of GPS data, is there anywhere that hosts orthophotos in a way that makes a 139Mb JPEG orthophoto actually usable by the general public?
Sample models:-
https://sketchfab.com/models/c0dc76861241475b8e8ecaca56514c68
https://sketchfab.com/models/417253342c924739a60d6478b54f6c28
Looking forward to the replies.
The results are very impressive, and worthy of publication, but both Mpabox and 4DMapper (the two API-ready hosts that come with Photoscan Pro) are really geared around GPS/mapping.
Mixing underwater and GPS is possible (and we are working on it) but in absence of GPS data, is there anywhere that hosts orthophotos in a way that makes a 139Mb JPEG orthophoto actually usable by the general public?
Sample models:-
https://sketchfab.com/models/c0dc76861241475b8e8ecaca56514c68
https://sketchfab.com/models/417253342c924739a60d6478b54f6c28
Looking forward to the replies.
56
General / Re: underwater photos don't align
« on: March 06, 2016, 12:07:33 PM »The photos are reasonably sharp and in focus (as far as underwater photos go…). I am using a point in shoot camera with an automatic ISO setting (it varies I believe based on the lighting conditions).
As others have said, overlap is critical. And a well exposed & lit image will be as sharp as anything taken on the land...
I have just run some material off of a Gopro video through Photoscan. The biggest issues was a lack of sharpness (low light/slow shutter/fast panning) in some frames resulting in a lack of images that overlapped and were technically acceptable.
Do you have a subset of the images that do align? If yes, then perhaps have a look at those and compare them to what does not?
Without overlap there is nothing to be recovered, so another dive is probably in order.
57
General / Re: underwater photos don't align
« on: February 29, 2016, 02:35:06 PM »
The camera and lens combination I use underwater isn't calibrated either, but the resulting images do align.
So I do not think the issue is with calibration.
Are you applying any post-shoot manipulation to the images?
What ISO are you working at when shooting the images?
Are the images in sharp focus?
So I do not think the issue is with calibration.
Are you applying any post-shoot manipulation to the images?
What ISO are you working at when shooting the images?
Are the images in sharp focus?
58
Feature Requests / Size of Marker ID text
« on: November 30, 2015, 07:16:34 PM »
When using Photoscan Professional -> Tools -> Markers -> Print Markers.
Relative to each printed Marker there is a number that can be used to ID the marker. Relative to the printed marker, the printed number is quite small.
Would it be possible to allow users to set a preference that controls the size of the text printed alongside the marker?
Relative to each printed Marker there is a number that can be used to ID the marker. Relative to the printed marker, the printed number is quite small.
Would it be possible to allow users to set a preference that controls the size of the text printed alongside the marker?
59
General / Re: Adding thickness to model
« on: September 09, 2015, 10:49:31 AM »
Thanks to all who replied on & off the forum.
Model is now thickened and printing...meshmixer did the job.
Model is now thickened and printing...meshmixer did the job.
60
General / Adding thickness to model
« on: September 04, 2015, 10:51:53 AM »
We are now printing in plastic the first 3D models of underwater artefacts and shipwreck remains, and the exported STL files from Photoscan are coming out very well indeed.
But the actual printing process can struggle with models with a greater level of texture/detail, and its necessary to build a printed frame around the model to help it build. Here's an example:-
https://www.facebook.com/theshipwreckproject/photos/a.128279717253739.33108.127175257364185/885165258231844/?type=1&theater
And the model itself:-
https://sketchfab.com/models/9d3ff4db7fca4471a2cd5e50ce4ee237
So ideally what I would like to do is add thickness to the surface in one direction only (the seabed side in this case) that will help the model build.
Has anyone found a tool or application that will add thickness to a surface? This is more in the realm of 3D CAD perhaps, but hoping someone here might be able to help?
But the actual printing process can struggle with models with a greater level of texture/detail, and its necessary to build a printed frame around the model to help it build. Here's an example:-
https://www.facebook.com/theshipwreckproject/photos/a.128279717253739.33108.127175257364185/885165258231844/?type=1&theater
And the model itself:-
https://sketchfab.com/models/9d3ff4db7fca4471a2cd5e50ce4ee237
So ideally what I would like to do is add thickness to the surface in one direction only (the seabed side in this case) that will help the model build.
Has anyone found a tool or application that will add thickness to a surface? This is more in the realm of 3D CAD perhaps, but hoping someone here might be able to help?
