Eliminating Radial Distortion

dgriffiths · Jun 24, 2017

Hi,

I have been mapping some very flat salt marshes using the P4P and getting a very obvious syncline in my SfM derived point clouds due to radial distortion. I have been reading up on this and it seems that parallel nadir flights paths (i.e. a standard GSPro mapping project) are the worst flight paths for experiencing this.

Has anyone on here got any experience using oblique imagery to reduce radial distortion, or perhaps use convergent flight paths? I will be heading back up on Wednesday to survey again and would be good if anyone had some tried and testing methods for this.

Dave

r62ewa · Jun 25, 2017

I don't know if it will help but have you tried the Double Grid mission in Pix4DCapture? It will fly the first traditional grid of parallel lines for the first grid. The second grid will be flown rotated 90 degrees to the first grid. When I used this feature previously I had the first grid flown with camera straight down and the second grid with the camera at 45 degrees to capture oblique imagery.

dgriffiths · Jun 26, 2017

r62ewa said:
I don't know if it will help but have you tried the Double Grid mission in Pix4DCapture? It will fly the first traditional grid of parallel lines for the first grid. The second grid will be flown rotated 90 degrees to the first grid. When I used this feature previously I had the first grid flown with camera straight down and the second grid with the camera at 45 degrees to capture oblique imagery.

Cheers for that, I have never used Pix4D but might check it out as that sounds like what I need. Have you tried it and compared results at all?

r62ewa · Jun 26, 2017

I have not compared results but the mission where I used the double grid mentioned above there were over 700 images taken and the result was very good. All I did was some linear measurement ground truthing which was very accurate.