01-12-2018 04:51 AM
I have a question if you can help me. I am working on ZY3 satellite images to extract a DEM. I have depended on the RFM to solve and extract the DEM using the RPCs files attached with the images. My question is about the elevations in this extracted DEM, are these elevations ellipsoidal or they are orthometric?. According to what i know, the Rational Functional Model (RFM) constructs the relation between (Lat., Long., and ellipsoidal heights) and (samples and lines) based on the values of the RPCs. So, I believe that these elevations of the extracted DEM are ellipsoidal elevations but I need to make sure about that. By the way, I did not use any control points. I just used RPCs files and automatic tie points.
01-12-2018 05:06 AM
I just wanted to make sure about that as i have read some papers on the same topic and they said that the elevations are orthometric not ellipsoidal. So, i was thinking maybe the Imagine Photogrammetry depends on using a geoid model like EGM96 or EGM2008 to convert to orthometric heights after generating the elevations.
01-12-2018 06:09 AM
"i was thinking maybe the Imagine Photogrammetry depends on using a geoid model like EGM96 or EGM2008 to convert to orthometric heights after generating the elevations."
It can do it but you must separately tell that to system when creating the block and obviously you have not done that. So Imagine can work with orthometric heights directly in block and make the conversions you refer but then all reference information must be on that height system and also you must tell that in block projection settings. I also assume that you must then have some ground control as RPC's will probably never be orthometric and you must manually feed some orthometric height data via GCP's.
If your block is based on RPC and tiepoints it can not ever be orthometric
01-12-2018 06:23 AM
I totally agree with what you said and what SamMegenta said. I know that RFM will lead to ellipsoidal heights and also know that these RPCs are bias. Actually some papers confused me about that as they mentioned that they didn't use any GCPs and no geoid model and nevertheless they said that they obtained orthometric heights this made me confused.
01-12-2018 06:34 AM
There is so much confusion in the world related on height measurement that I am not suprised. The only thing we can be sure is that height of some place depends where the zero levels happened to be set. So height varies a lot in case by case and telling absolute truth about it is always very hard and easy to mess with. It is easy to say some building is higher than another in your visible neighbourhood but things gets much trickier if you compare mountain in North-America to some other montain in South-America. What is the absolute height of some mountain that is standing on ground of nearly ellipsoid formed but still very vague ball and basis of each mountain (continents) are floating over melted rock. It is suprising that concept of height is already working as well as it currently does.
Height is a tough question - that is only thing we can be sure at