So in lieu of orthorectification (the drone I'm using does not provide geographic data) I have decided to try georeferencing as a means of overlaying aerial photography onto my DEM. Before I go out there and start taking some control points with the differential GPS I was hoping I could get some advice. I've done this in class before and if I remember correctly 6-10 control points is optimal per image, correct? Secondly, I am doing this for an archaeological site on a mountain the Andes. So I can survey the actual units for control points, but other than that, there's not much else I can use as control points in some areas of the mountain. There's cacti here and there, but nothing that really stands out. In a classroom environment I've done this with urban maps where buildings and road intersections could be used. This is the desert and pretty barren. Also, if someone could explain the process of georeferencing to me a little better I'd appreciate it. I understand labeling control points and what not and giving the image a projection to match the DEM. But the end result should be a mosaic of the images I've gathered, correct? Thanks again for your help!
I'm starting to piece this together. Every time georeference one image onto some control points it's visible until I put another one on there. What I'm going for is a mosaic. Like I said, it works until I georeference the next image, then the most recent one is the only one visible. What am I doing or not doing? Thanks
I'm also having trouble with the images. The images are gigantic compared to the DEM. I zoom in on the DEM, but then the images does the same thing, and this is even before i start adding in the control points. Is there anyway I can resize the images for creating a mosaic. Should I be saving the images in different formats (they're in .jpg)?
You need orthorectify first, if you not orthorectify the images, your images have distortions and this not represents the reality.
For this you can use a Photogrammetry Suite Software.
Unfortunately I can't orthorectify. The imagery I have is from the drone we are working with and it only takes photos, it does not provide geographic information. As a result, the raster dataset does not have RPCs and I am unable to orthorectify. Georeferencing is my only option. Any suggestions?
I can talk you through a few of your options if you can tell me more about your data and what you hope to achieve.
I hope that is helpful - note we just did several workshops at the Esri User Conference that may be helpful to you - we can send a PDF document re: Block Adjustment tools and another re: georeferencing drone imagery using the Frame Camera Raster Type
Do you have update on this post of yours. You have mentioned -
"a new tool coming from Esri later this fall for georeferencing drone imagery (check back with me in a few months)"