deformed buildings in Photogrammetry
-
- V.I.P Member
- Posts: 846
- Joined: Tue Feb 26, 2013 9:32 pm
- 11
- Full Name: Massimo De Marchi
- Company Details: Massimo De Marchi
- Company Position Title: freelancer
- Country: Italy
- Skype Name: massimo_de_marchi
- Has thanked: 15 times
- Been thanked: 53 times
- Contact:
deformed buildings in Photogrammetry
I would like your opinion about photogrammetry software.
By the way.. would it be possible to add a forum sections?
I did a phantom 4 scan of a set of buildings, 3 cm GSD (about 65-70 meters from ground).
I did a test with photoscan personal, Pix4d and reality capture demo, but the buildings are inclined on one side.
The buildings are about 15 meters high. The effect is systematic, and I cannot figure why it happens and how to fix it.
Any insight or suggestion is welcome.
Massimo
By the way.. would it be possible to add a forum sections?
I did a phantom 4 scan of a set of buildings, 3 cm GSD (about 65-70 meters from ground).
I did a test with photoscan personal, Pix4d and reality capture demo, but the buildings are inclined on one side.
The buildings are about 15 meters high. The effect is systematic, and I cannot figure why it happens and how to fix it.
Any insight or suggestion is welcome.
Massimo
You do not have the required permissions to view the files attached to this post.
ing. Massimo De Marchi - +39 347 32 17 049 - www.studiodemarchi.net
- pedropcsena
- I have made 60-70 posts
- Posts: 62
- Joined: Wed Jul 27, 2016 2:08 pm
- 7
- Full Name: Pedro S.
- Company Details: I.B.W.
- Company Position Title: LiDAR-3D Laser Scanning Specialist
- Country: Canada
- Linkedin Profile: Yes
Re: deformed buildings in Photogrammetry
Maybe it is due to the marker and camera position accuracy that you set?
If you set the values too low, the software tries to make your parameters fit therefore causing distortion.
When I worked with Photoscan I set my camera position accuracy to 10 meters and the ground marker to 0.05, even though my ground control network was way better than that - I set it with total station and performed a least squares adjustment.
Cheers,
Pedro
If you set the values too low, the software tries to make your parameters fit therefore causing distortion.
When I worked with Photoscan I set my camera position accuracy to 10 meters and the ground marker to 0.05, even though my ground control network was way better than that - I set it with total station and performed a least squares adjustment.
Cheers,
Pedro
-
- V.I.P Member
- Posts: 846
- Joined: Tue Feb 26, 2013 9:32 pm
- 11
- Full Name: Massimo De Marchi
- Company Details: Massimo De Marchi
- Company Position Title: freelancer
- Country: Italy
- Skype Name: massimo_de_marchi
- Has thanked: 15 times
- Been thanked: 53 times
- Contact:
Re: deformed buildings in Photogrammetry
Thanks for the reply.
This is before inserting the GCP (with photomodeler basic GCP are not available by the way).
I'm not sure I would have any advantage when using them, mainly because my set of GPS surveyed points lies at floor level.
I'm checking with Altizure too..
Massimo
This is before inserting the GCP (with photomodeler basic GCP are not available by the way).
I'm not sure I would have any advantage when using them, mainly because my set of GPS surveyed points lies at floor level.
I'm checking with Altizure too..
Massimo
ing. Massimo De Marchi - +39 347 32 17 049 - www.studiodemarchi.net
-
- V.I.P Member
- Posts: 904
- Joined: Fri Apr 20, 2012 1:49 pm
- 11
- Full Name: Thad Wester
- Company Details: Clarity Scanning
- Company Position Title: President
- Country: USA
- Linkedin Profile: Yes
- Location: SC
- Has thanked: 1 time
- Been thanked: 2 times
Re: deformed buildings in Photogrammetry
Any reason to believe the buildings are not slanted? the stacks of material on the ground do not look distorted.
Cant think of any other reason... Could try a quick trial of pix4d. I've had better luck with pix4d with drone data.
Cant think of any other reason... Could try a quick trial of pix4d. I've had better luck with pix4d with drone data.
-
- V.I.P Member
- Posts: 846
- Joined: Tue Feb 26, 2013 9:32 pm
- 11
- Full Name: Massimo De Marchi
- Company Details: Massimo De Marchi
- Company Position Title: freelancer
- Country: Italy
- Skype Name: massimo_de_marchi
- Has thanked: 15 times
- Been thanked: 53 times
- Contact:
Re: deformed buildings in Photogrammetry
Thanks.
The mid image is the Pix4D trial.
The last is reality capture trial.
The three show the same behaviour, more or less.
I'm confident the buildings are not slanted (This is just one of the many buildings, all slanted in the same way..).
It looks like distortion increases along z.
Massimo
The mid image is the Pix4D trial.
The last is reality capture trial.
The three show the same behaviour, more or less.
I'm confident the buildings are not slanted (This is just one of the many buildings, all slanted in the same way..).
It looks like distortion increases along z.
Massimo
ing. Massimo De Marchi - +39 347 32 17 049 - www.studiodemarchi.net
- pedropcsena
- I have made 60-70 posts
- Posts: 62
- Joined: Wed Jul 27, 2016 2:08 pm
- 7
- Full Name: Pedro S.
- Company Details: I.B.W.
- Company Position Title: LiDAR-3D Laser Scanning Specialist
- Country: Canada
- Linkedin Profile: Yes
Re: deformed buildings in Photogrammetry
I am not sure then... But as a heads up, when you bring the GPS coordinates you might have issues with the Z values. The software uses the ellipsoid height and not the geoid height. It is necessary to download the geoid file in order to have it working properly.
-
- V.I.P Member
- Posts: 846
- Joined: Tue Feb 26, 2013 9:32 pm
- 11
- Full Name: Massimo De Marchi
- Company Details: Massimo De Marchi
- Company Position Title: freelancer
- Country: Italy
- Skype Name: massimo_de_marchi
- Has thanked: 15 times
- Been thanked: 53 times
- Contact:
Re: deformed buildings in Photogrammetry
This is a strange issue..
photoscan basic is not using any king of geotagging nor GCP and is offering a slightly better result.
Altizure looks even better. I have to download the data and check.
For a 200 meters project the geoid-ellissoid offset can be assumed as constant, moreover I used GCP in Pix4D in local coordinates in the past, but no buildings. Anyway they worked fine.
Phantom used to place a relative height in the picture tag, now it's more like an ellipsoid.
I can transform my GPS measurements to sea level, but that cannot be the issue. It's just a constant offest for such a small patch.
Massimo
photoscan basic is not using any king of geotagging nor GCP and is offering a slightly better result.
Altizure looks even better. I have to download the data and check.
For a 200 meters project the geoid-ellissoid offset can be assumed as constant, moreover I used GCP in Pix4D in local coordinates in the past, but no buildings. Anyway they worked fine.
Phantom used to place a relative height in the picture tag, now it's more like an ellipsoid.
I can transform my GPS measurements to sea level, but that cannot be the issue. It's just a constant offest for such a small patch.
Massimo
ing. Massimo De Marchi - +39 347 32 17 049 - www.studiodemarchi.net
- 3DForensics
- Honorary Member
- Posts: 1979
- Joined: Mon Aug 03, 2009 1:52 am
- 14
- Full Name: Eugene Liscio
- Company Details: AI2-3D Forensics
- Company Position Title: Owner
- Skype Name: eliscio
- Location: Toronto, Canada
- Has thanked: 13 times
- Been thanked: 70 times
- Contact:
Re: deformed buildings in Photogrammetry
I am wondering if something about how you flew the flight path and the rolling shutter on the camera. Did you happen to have the camera pan sideways??
Sent from my iPad using Tapatalk
Sent from my iPad using Tapatalk
-
- V.I.P Member
- Posts: 846
- Joined: Tue Feb 26, 2013 9:32 pm
- 11
- Full Name: Massimo De Marchi
- Company Details: Massimo De Marchi
- Company Position Title: freelancer
- Country: Italy
- Skype Name: massimo_de_marchi
- Has thanked: 15 times
- Been thanked: 53 times
- Contact:
Re: deformed buildings in Photogrammetry
I was thinking about the rolling shutter issue, but I flew a grid mission, so going back and forth they should compensate, I would say. Moreover recent Pix4D versions should be able to compensate also the rolling shutter.
Altizure output is much better, by the way.. (no GCP).
I veryfied the orthophoto too, I had to scale it a little bit, but the result is quite nice and the deviation is reasonably limited.
Massimo
Altizure output is much better, by the way.. (no GCP).
I veryfied the orthophoto too, I had to scale it a little bit, but the result is quite nice and the deviation is reasonably limited.
Massimo
ing. Massimo De Marchi - +39 347 32 17 049 - www.studiodemarchi.net
-
- V.I.P Member
- Posts: 846
- Joined: Tue Feb 26, 2013 9:32 pm
- 11
- Full Name: Massimo De Marchi
- Company Details: Massimo De Marchi
- Company Position Title: freelancer
- Country: Italy
- Skype Name: massimo_de_marchi
- Has thanked: 15 times
- Been thanked: 53 times
- Contact:
Re: deformed buildings in Photogrammetry
A short update:
I tried some of the on line services, and I still have the same issue. The only notable exception is Altizure.
I flew the Altizure app, but as far as I know it's not reasonable to consider the possibility for the app to tamper with data (DJI sdk has many limitations). Maybe I'l cross check and try to fly the same mission with different apps, but I'm confident data logging cannot be modified. I'm also confident the building is not slanted, and in any case it cannot be THIS slanted.. (I laser scanned it, by the way)
I'm attaching some othophotos of the tallest building, using Altizure, Maps made easy, Drone deploy and reality capture.
Massimo
I tried some of the on line services, and I still have the same issue. The only notable exception is Altizure.
I flew the Altizure app, but as far as I know it's not reasonable to consider the possibility for the app to tamper with data (DJI sdk has many limitations). Maybe I'l cross check and try to fly the same mission with different apps, but I'm confident data logging cannot be modified. I'm also confident the building is not slanted, and in any case it cannot be THIS slanted.. (I laser scanned it, by the way)
I'm attaching some othophotos of the tallest building, using Altizure, Maps made easy, Drone deploy and reality capture.
Massimo
You do not have the required permissions to view the files attached to this post.
ing. Massimo De Marchi - +39 347 32 17 049 - www.studiodemarchi.net