Finally I have the first results which are good but also have some questions.
I did 2 flights, one at 60m and one at 40m. At first I had just 5 GCP's which finally wasn't enough for good verticals. The orthophoto came pretty accurate and the linework I had done matched within cm.
However to achieve tight verticals I had to go back and measure some more GCP's that was shown in the pictures like manholes, corners etc.
Finally with 20GCP's about 50m apart the accuracy was very good both horizontal and vertical.
In the 40m flight I had even better results from the 60m but had some issues also.
If i configure the software (Contextcapture for the moment) to adjust the Focal Length and compute the distortion of the camera it produces a solution that is good on horizontal RMSE but the verticals are not very good (had 20cm difference in a check point and some other points I had measured with GPS-RTK). The Focal Length was computed at 3.30mm instead of 3.61mm which is the EXIF value of the photos. However if I choose to fix the Focal Length at 3.61mm the software produces a solution that is worse in horizontal RMSE at about 3-4cm BUT an impressive vertical RMSE of 1-2cm which is confirmed in the check points I had measured! This is odd because the horizontal accuracy supposed to be better that vertical.
My question is, should I keep the Focal Length fixed in 3.61mm? I don't understand exactly what is going on with focal length. For example a flight of 60m vs a flight of 40m altitude has the same focal length? Because I am using terrain awareness feature not all the photos have the same flight altitude, so is the focal length the same on all photos? Sorry for my ignorance in Photogrammetry but as a surveyor I took a class of it 15 years ago and didn't bother with it until now.
I did 2 flights, one at 60m and one at 40m. At first I had just 5 GCP's which finally wasn't enough for good verticals. The orthophoto came pretty accurate and the linework I had done matched within cm.
However to achieve tight verticals I had to go back and measure some more GCP's that was shown in the pictures like manholes, corners etc.
Finally with 20GCP's about 50m apart the accuracy was very good both horizontal and vertical.
In the 40m flight I had even better results from the 60m but had some issues also.
If i configure the software (Contextcapture for the moment) to adjust the Focal Length and compute the distortion of the camera it produces a solution that is good on horizontal RMSE but the verticals are not very good (had 20cm difference in a check point and some other points I had measured with GPS-RTK). The Focal Length was computed at 3.30mm instead of 3.61mm which is the EXIF value of the photos. However if I choose to fix the Focal Length at 3.61mm the software produces a solution that is worse in horizontal RMSE at about 3-4cm BUT an impressive vertical RMSE of 1-2cm which is confirmed in the check points I had measured! This is odd because the horizontal accuracy supposed to be better that vertical.
My question is, should I keep the Focal Length fixed in 3.61mm? I don't understand exactly what is going on with focal length. For example a flight of 60m vs a flight of 40m altitude has the same focal length? Because I am using terrain awareness feature not all the photos have the same flight altitude, so is the focal length the same on all photos? Sorry for my ignorance in Photogrammetry but as a surveyor I took a class of it 15 years ago and didn't bother with it until now.
