Messages

76

Bug Reports / DEM export error from different ellipsoid

« on: July 17, 2017, 09:45:10 AM »

Hi Alexery,

I found Agisoft export DEM from other ellipsoid like Airy 1930 or AGD66 to WGS84, the DEM will be shifted a lot. For example, if the DEM is on AGD84 with elevation of 110, if I directly export the DEM into WGS84, the elevation will be -35m.

Hope this would be fixed or identified if I did something wrong.

Thanks,

77

General / Re: DEM error but GCP RMS is very small

« on: July 13, 2017, 09:08:13 AM »

Hello Alexey,

I mean in the GCP it has elevation of 94.366ft, and the alt error is 0.03ft, but the estimated elevation is 93.466ft, which is 0.1ft difference. How can I force the model to use GPC's height? I also noticed that the photos have slightly curvature at the edge of the flight area after optimising the cameras.

Thanks, 

78

General / Re: DEM error but GCP RMS is very small

« on: July 13, 2017, 08:49:37 AM »

Hi Alexey,

I got another dataset which is also wired, the photos flew along a road, and before importing GCPs the altitude error is about 2m, but after imported GCPs and reoptimise, the easting/northing/alt error all get at about 10cm at the GCPs and 11m at photo alt error. I checked the photo altitude should match the GCP elevation (before reoptimising).

I tried to set the marker accuracy to 0.001 and the RMS shows smaller at GCPs, but the estimated GCP elevation is much away to the actual GCP elevation.

Could you help to check what's wrong for the problem? I'm trying 1.3 and the photos are from DJI model FC6310.

Thank you.

79

General / Re: DEM error but GCP RMS is very small

« on: July 09, 2017, 12:08:17 AM »

Hi Alexey,

In the alignment step I used default 40,000 points settings, and in the dense point cloud step I used medium quality, and for DEM I used Dense point cloud and enabled interpolation.

Would these parameters help?

80

General / Re: DEM error but GCP RMS is very small

« on: July 05, 2017, 07:38:42 AM »

Hi Alexey,

Sorry to raise this issue again, I recently met a dataset with RMS within 1 cm (Due to I set the marker accuracy as 0.002, otherwise the RMS have two points at 0.1m). So I forced the model with the GCPs, but the final result shows the elevations shifted at 0.4m or more for each GCPs. Is there possible be some reasons?

Thanks a lot!

81

Bug Reports / Re: CGVD2013 reports "vertical datum out of range"

« on: June 26, 2017, 08:17:23 AM »

Hello Alexey,

Thanks for the reply, sorry for the late update, I restarted PhotoScan and it recognise the files now! Thanks, this works now!

Thank you!

82

Bug Reports / CGVD2013 reports "vertical datum out of range"

« on: June 22, 2017, 06:51:17 AM »

Hi Alexery,

I tried to use the CGVD2013 geoid file, but it reports "vertical datum out of range". I checked the photos and I opened the geotiff they both are in the correct position, is there something wrong in Agisoft 1.3? Also I found NAVD88 12 a and 12 b can be read in, however this geoTIF does not have projection / datum defined, would this be a correct file?

Thanks a lot!

83

General / Re: DJI FLIR camera Pixel size not calculated

« on: June 20, 2017, 10:57:35 AM »

Hi Alexery,

Thank you! This works! Thanks a lot, so the sensor size should be put into the pixel size. Thanks a lot!

Cheers,

84

General / Re: DJI FLIR camera Pixel size not calculated

« on: June 19, 2017, 08:40:05 AM »

Hi Alexey,

Thank you for your kind reply, I just searched and got some parameters of DJI FLIR camera, but using the value doesn't allow me to align the photos correct. As in the camera calibration page, the Pixel Size (mm) is showing "unknown", I tried to input the value: Camera sensor size is 17 micron meter, while photo resolution is 780 * 480. so I used 0.017 / 480 but got a very small number and the cameras didn't aligned correctly.

Could you tell me if there's something wrong in the calculation of the pixel size (mm)?

Thanks a lot!

85

General / DJI FLIR camera Pixel size not calculated

« on: June 05, 2017, 11:58:12 AM »

Hi Alexey,

Is it possible that DJI FLIR camera's Pixel size in camera calibration can be extracted? How could we manually calculate it? Currently when I load the FLIR photos in the camera calibration will show "unknown" at the pixel size (mm). The only information shown is the focal length of 13mm.

Thanks a lot.

86

Python and Java API / Re: Tie point management

« on: May 19, 2017, 04:33:55 AM »

Hello Alexey,

Thank you for your quick response! However, what I did is manually created a tie point (e.g. Point 1), then how can I get the point.coord for it? It is not showing in the marker.reference.

Thanks a lot!

Cheers,

87

Python and Java API / Re: Tie point management

« on: May 18, 2017, 10:12:51 AM »

Hi Alexey,

So is it possible to get the tie point estimated location like marker.reference.location? Is there a tiepoint.location function in 1.3?

Thanks a lot.

88

Bug Reports / DEM export wrong by directly changing coordinate system

« on: March 29, 2017, 04:03:05 AM »

Hi Alexey,

We found when exporting DEM in geoTIF generated in OSGB or any coordinate systems which are not based on GRS80, the height doesn't match with it was generated in WGS84 coordinates. For example, I generate a DEM in OSGB epsg 27700, which is based on Airy ellipsoid, and direct export, and when exporting I select the coordinate system as "WGS84", so it should transform to WGS84, and horizontally it does correctly, however, the height is shifted by about 3.8 meters.
See image attached.

This can be corrected by regenerating DEM in WGS84 coordinate system defined, and export would be correct. Would it be a bug or it is not suggested to export in different coordinate system in Agisoft?

Thanks,

Cheers,

89

General / Re: DEM error but GCP RMS is very small

« on: March 28, 2017, 12:34:04 PM »

Hi Alexery,

How can I give you the photos?

Chees,
Tim

90

General / Re: DEM error but GCP RMS is very small

« on: March 22, 2017, 02:51:30 PM »

Hello,

Thank you for your reply! The camera parameters I used are all the parameters in reoptimising, and in DEM generation I use dense point cloud and other parameters used default.

Thanks!