Topics

16

General / Creating line drawings of models - which software to use?

« on: August 30, 2014, 01:08:42 PM »

Hi everyone!

I'm looking for a good workflow where I can export models from Agisoft PhotoScan, and then trace relevant features (in the geometry, but preferably also on the texture) in order to output a lines drawing containing only the relevant details.
The actual case study will be a shipwreck site, where I want to trace both a top view and cross sections of the wreck for archaeological purposes.

Available software (which I think might be useful): 
- Rhinoceros 4.0
- Illustrator
- Quantum GIS

For the top view I suppose I could just export an orthophoto of the entire site and trace that in Illustrator (unless someone has better suggestions?) but for the cross sections I was thinking of some approach where I imported the model in Rhino, dissected the model to the cross-section I wanted to draw, and then traced that. However Rhino doesn't appear to be very good at "reading" or showing the texture generated by PhotoScan (ie the texture is always a lot less good in Rhino), which is an issue.

I would be very grateful to hear from anyone who has had some experience with similar issues!

Thanks in advance,

Thomas

17

General / Can you view two unaligned chunks at once?

« on: August 06, 2014, 06:35:35 PM »

Hey everyone!

I was wondering whether it's possible to have two models from the same project (for instance two chunks) open each in their own "model" pane within one PhotoScan project?

The reason I would like to do this is to be able to manually add markers to overlapping areas between the chunks, while viewing both chunks at once - in order to eventually perform marker-based chunk alignment. At the moment when I want to do this I have to switch back and forth between the chunks, they're each in their own coordinate system so I usually have to rotate the chunks again to find the right spot for the marker, and it just makes it very time-consuming and complicated.

Please let me know if you know of a way to view two chunks at once within one project!
Thanks,

Tom

18

General / 360° camera - PhotoScan compatibility?

« on: June 20, 2014, 08:05:07 PM »

I wonder how compatible Agisoft PhotoScan would be with a camera like this? https://www.kickstarter.com/projects/giroptic/the-worlds-first-full-hd-360-camera?ref=nav_search

19

Feature Requests / Run photo align -> multiple chunks detected

« on: June 16, 2014, 01:51:34 PM »

Hi Agisoft,

I'm doing a lot of work with rather poor data sets, like legacy images taken many years ago, not necessarily ideal for photogrammetry but trying to get the best results I can out of them.

Using this type of data what I often see is that I try to align for instance 100 pictures, and then PhotoScan only manages to align 30 or so of them (even though the pictures do in theory overlap -> probably issues with image quality). However, if I then take the remaining 70 pictures into a new chunk, PhotoScan will again manage to align say 25 pictures, creating another big chunk of the subject I'm trying to model.

It would be nice if instead of having to re-run the photo alignment phase each time, Agisoft PhotoScan would just say something like "PhotoScan has detected multiple chunks in this data set - save to different chunks?". And then save the 30 first aligned images into the initial chunk, and the 25 'secondary' aligned images into a new chunk.

I know VisualSfM has a feature like this (when you use up or down arrows you switch between the different chunks it's found) so I'm sure it's not too hard to implement

Tom

20

General / Manual photo alignment - tips?

« on: June 16, 2014, 01:42:26 PM »

Hey everyone,

I've been working on a dataset of 58 pictures, taken with two different analogue cameras in the year 2000. The object of study is an ancient ship, the Belgian "Kogge" (http://kogge.be/). There's no 3D model of the site and it would be very cool to make one 14 years after excavation.
Since the pictures have been scanned from dias and hard copies there's no exif data, and some of the pictures are rather poor quality. Additionally they've been taken on different dates, and so objects around the vessel change from day to day. Unfortunately this is the data we've got to work with so here's my workflow:

- masked the pictures so they include only the vessel (ie avoid changing objects around it)
- automatic photo align at high accuracy, pair preselection generic, 40k point limit, constrain features by mask
=> after this phase 23/58 pictures were aligned, covering the starboard side of the vessel (see attachment)
- manually photo align by manually putting markers on unaligned as well as aligned pictures, and one by one choosing "align selected cameras" for unaligned photos.
=> after adding roughly 120 manual markers in this way I've aligned 42/58 pictures, covering both sides of the vessel (see attachment 2)

Now, while these manually aligned pictures do in fact cover every part of the port side, there's still massive holes in the dense reconstruction. Does anyone have any ideas on how to improve this result? I feel like PhotoScan only uses the matches it's found in the initial automatic "Align Photos" phase to make the sparse point cloud. Is there any way to sort of tell the software "well now that I've manually shown you where this picture goes, look really thoroughly around that area for matches in nearby pictures"?

Anyone have experience with manual picture alignment?

Thanks in advance for any advice you may have!
Cheers,

Tom

21

Feature Requests / "Add Videos" feature

« on: May 27, 2014, 12:31:31 PM »

Hi Agisoft!

I've been playing around with Photoscan for a couple of months now and I think a really cool feature to add would be an "Add Videos" button (like, right next to the "Add Photos" button). I'm sure lots of others have already considered this so please +1 if you agree with my arguments and think this should become an awesome Photoscan feature.

Advantages of video input would include:
- very dense overlaps between frames
- reduced recording times
- often a larger depth of field compared to pictures
- ... (maybe other people can think of other advantages?)

Workflow could be:
- click "Add Video" button, select video and add it to Photoscan
- Photoscan extracts frames of just one second of video footage, in order to analyse frame rate and resolution + matches points within that one second (so say 30 frames) of footage in order to determine quality for Photogrammetry. If after analyzing, the footage is deemed unsuitable for Photogrammetry, Photoscan alerts you that you will probably get poor results -> "do you wish to continue anyway?".
- based on the above analysis Photoscan offers you several quality options from very low (bare minimum of frames extracted) to very high (based on the analysis Photoscan knows how many frames it should extract for a good model) to ultra ridiculously high (every single frame extracted just for overkill purposes).
-> remaining steps are the same as with pictures.

Now, before people start hitting me with a stick for suggesting video rather than picture input, consider the image in the attachment. It's a model of a shipwreck site generated based on 44 images extracted from a 2:11 minute unstructured video recording (as you can see from the camera positions) with a GoPro HD2 underwater.
So no EXIF data + fisheye lens + unstructured recording + underwater distortion = still a very good model. And this has worked time and time again on multiple projects

Regards,

Tom

22

General / 1980s Shipwreck Excavation - help with camera calibration?

« on: May 23, 2014, 10:32:38 PM »

Hi Agisoft users!

In the 1980s and early 90s Dutch archaeologists attempted to record the wreck of a 17th century trading ship (the Aanloop Molengat site) with photogrammetry. However due to the technical limitations of the time and some problems with the captured pictures, photogrammetry never really worked out.

Using Agisoft Photoscan I now hope to reuse the original pictures to finally create a 3D model of the site. However the pictures were obviously captured analogue, printed and then scanned digitally some years later (-> the data I have now) so no EXIF data is available. The alignment results are quite bad for the time being, but by adding the camera parameters I hope to improve alignment results.

What I know is:
- the pictures were taken in stereopairs
- the cameras used are described as "Hasselblad / Ocean Optics MC-70"
- the camera had a 47mm wide angle lens, covering 94° diagonally underwater
- the film used was a "70 x 60mm format slide positive film"
- the operating distance was set to a maximum of 1.5m due to lack of visibility and light

This info comes mainly from the caption beneath the first picture of this article: https://www.academia.edu/925308/Archaeological_Mapping_A_Sub-North_Sea_Experiment

Does anyone have any idea how/whether I can translate this info into useful camera calibration settings for Agisoft Photoscan?
It would be really cool to generate a 3D model of an archaeological site using this type of old "legacy data" so I would be really grateful for any advice you may have on the topic!!

Sincerely,

Thomas Van Damme
Maritime Archaeology student at the University of Southern Denmark



(update) ADDITIONAL INFO:

- Aanloop Molengat site plan: https://easy.dans.knaw.nl/ui/resources/easy/content?sid=dans-jumpoff:6256&did=JOF_veldtekening_scheepswrak_A4.jpg

- Sample of a stereopair (after scanning of the original analogue pictures): http://www.mediafire.com/view/4hhfps53i5do1g6/102-003N.jpg and http://www.mediafire.com/view/4buo6014e2svayz/102-003Z.jpg

23

General / Chunk alignment options - boat model

« on: May 03, 2014, 11:57:07 AM »

Hi Agisoft Photoscan users!

I have a dilemma. I have generated a nice model of a historic sailing vessel, but I have generated it in 4 chunks: bow, stern, middle hull and sails. Now chunk alignment is finding the approximate positions, but it's not entirely perfect.

In order to combine the chunks accurately, and get the best possible result should I:
- export the 4 point clouds, align in meshlab, somehow import the aligned clouds back into Photoscan (will that still allow me to texturize them in Photoscan?)
- make a mesh of each individual chunk, tools -> import masks for each chunk, throw the masked photos into one chunk, and reprocess everything?
- not mask the individual chunks and just realign all pictures?

The reason I have processed in chunks is because there are several areas of the boat that look very alike (like stern and bow) but actually aren't. I'm afraid that if I process everything together, they will be misaligned (I had this previously on a boat that was very symmetrical, where Photoscan considered the starboard and port to be the same => it thought all pictures were taken from port rather than from both sides). Also, processing all together will take more computer time.

My thanks for any advice you may have!

Tom