Hi All

I have this problem a weird vertical shift, it comes down to the use of EGM96 and WGS84 datums. The problem is that, the areas was flown using two drones, but one drone produces GPS log using EGM96 the other produces GPS log just using WGS84 datum, as a result when the data is processed the vertical shift is observed clearly in areas surveyed by one drone as comparison to areas surveyed by the other. My question is that, how do we ensure then that data is processed using just one datum even if data is collected using different datums? This can not be change in the drones GPS so it will continue to be a problem if there is no way of fixing t during processing.

Any help will be highly appreciated, I have also sent a ticket to support.

Clarence

6 comments

• Hello Clarence,

It is true that you can only select one image geolocation coordinate system for the entire project, so if you import all the images in the same project, you can either select WGS 84 or egm 96. Also, you can only select one output coordinate system.

We would recommend you to convert the coordinates of the flight with the WGS 84 coordinates to the egm 96 vertical coordinate system. To do so, you can just add the N value (difference between the ellipsoid WGS84 and the geoid 96; varies between 30-100 meters depending on the location). If you know this difference, you can just add it to the Z values of the image geolocation, import all the images in the software and select egm 96.

If you do not know this value, you can easily compute it based on your datasets. The N difference between the two surfaces is unique for an area as big as a drone can cover. You can process part of the the project that has image geolocation in WGS 84 and check the altitude of one point on the ground. The vertical coordinate system of the model will be WGS84 and there will be a shift with the expected Mean Sea Level value of this point (which you can find it in Google Maps). This is the N value that will allow you the conversion between the two vertical coordinate systems.

Once you have converted the coordinates, you can add all the images in the same project, select egm 96, the desired output coordinate system and process the project as per normal.

Best regards,

• I sometimes get a large drop off on the edge of my dsms right at the edge of the processing area. They plummet to 0. Is this possibly related to the fact that I'm using WGS84 photography and exporting to state plane? I'm not choosing egm 96 in my image coordinate settings options or my output coordinate options.

Thanks,

• Hello Bill,

Is the area at the edge of the DSM covered by images or does it consist of big triangles generated after triangulating the edges of the point cloud? Could you please send us a screenshot of the area you are referring to?

Best regards,

• It's been more than a month since I wrote this so we've had some time to work on this.

My client is seeing this anomaly in Civil3d. See attached screen shots. He is creating a "processing area" around the section of interest, processing levels 2-3 then importing the dsm into civil 3d. The area dropping to "0" altitude is the area that is not included in the new "processing area". There is data there in the level 1 "Tie Points" point cloud. He is expecting to see a clean edge around the processing area not a cliff down to "0".

What is odd is that he does not always get this result and we are trying to determine why.

Thanks,

• Hello Bil,

Have you tried to import the DSM in another software, like Global Mapper (extended free version)? Do you face the same issue?

Could it be that the software you are using (Civil 3D) interprets the pixels around the model as if they had zero value? The DSM will be framed in a box like in the screenshot attached. The empty pixels around the model will have a value that corresponds to void and should be interpreted as such by the visualization software. Could it be that Civil 3D interprets the void pixels as pixels of value 0?

The DSM will be framed in a box like in the screenshot attached below. The empty pixels around the model will have a value that corresponds to void and should be interpreted as such by the visualization software. Could it be that Civil 3D interprets the void pixels as pixels of value 0?

Best regards,

• Global Mapper does not display data outside the box around the model so I think it is just how Civil3D displays. The client is looking to see if he can adjust that feature in his civil3d.

Thanks!

Bill

Please sign in to leave a comment.