Hi everyone,
I am new user to ContextCupture and I am working on a pilot project for a Telecommunication company in my country. We use a DJI Mavic Pro UAV for the video capture of a telecommunication antenna at 3840x2160 dpi in D-Cinelike mode. The final capture is a 336 sec video around the antenna tower following a helix way I tried several times to do the AT and the reconstruction of the antenna. However, the produced result is part of the around scenery or something abstract.I believe that the problem has to do with the sensor size and the focal length numbers set in CC. According to UAV specs these numbers should be respectivly 7,66 mm and 28 mm but I am not sure.
We are a little bit lost here and some help would be appreciated. Thanks.
At first you shouldn't use video at all. Video has motion blur which is not calculated in Contextcapture and you don't have GPS positions. Also don't use 60fps setting. Secondly if you use video then at least do one photo before video with same settings/zoom as for video to get correct metadata from this photo. For towers there is now special setting block type:orbit which helps because otherwise background takes over. Overall good luck but you may have to fly again taking photos.
MAVIC PRO
Sensor size:6.16mm
35mm: 4.73mm
Hi Bill,
I own a Mavic Pro too, and I don't do video for this kind of mission at all... I take only pictures.
While on site, I try to take as many pictures as I can with different flight parameters (distances, heights, camera angle,etc), so I don't have to come back again.
So, I don't think the problem is about CC settings, but about the camera parameters and flights planning.
DELORD Alicia
Also you may found these information helpful for your project.
https://www.youtube.com/watch?v=Wg4EhUeXHoI&index=23&list=PLMxDUk91MUDTYI8Uc70w1BWqfU9nkpTsJ&t=0s
https://bentley.sharefile.com/d-s1c825681c184d9ab
Thank you all for the advises. We will try again to reconstruct the model following your guidelines. I appreciate your help.
As Oto and Alicia said, videos are not suitable in this case of use. Photographs with the higher resolution you can are mandatory. Models representing thin structures are not easy to compute.
Also, if you are using the latest version of CC (4.4.9.516), you have the ability to modify the block type. Before any AT, click on the "Block_x", than choose the tab "Additionnal data" and select the block type "Orbit around thin vertical structure". This can help you to make model better.
HTH
Philippe HALET
CEO at beCAD a Gold Channel Partner