01-20-2018 07:43 PM
I am going to work a lot of aerial photos I have on my external hard drive. They do not have georeferenced in them. It was scanned by a contractor and they were flown over one of the National Forest area in Central Colorado, USA.
I have several questions for you out there that I may have trouble with.
For the input I have them as a jpg but I export them to TIFF files.
For the Reference image I use the NAIP 2015 files that are in MrSID. They are really huge area and are covered in by county.
For the Z values, I have a DEM that are in 10 meters.
For the aerial photos They are very old and are in black and white,
I have the instruction on how to perform them that I had from the training materials. I have been reading and research them. I still have not grasp about this.
So when I set them up and I have a question about the project properties, my question to you for the Geometric Model, which one would be correct for the aerial photos ?
For the Reference image, do they need to be bigger area than the input image ?
how many tie points do you recommend for this ?
Do they need to have the same pixel size for both the input and Reference ?
Let me know if you have any question you might have.
Solved! Go to Solution.
01-24-2018 08:41 AM
Have you reviewed the IMAGINE Online Help ‘AutoSync User Guide’? It addresses all of the topics that you are asking about.
Help > User Guides > AutoSync User Guide.
If the input is unrectified and the reference image is orthorectfied, you will want to use a 3D geometric model with an accurate DEM. Otherwise there may be issues with vertical terrain displacement. In your case using Direct Linear Transform would be an option.
Reference image area
Ideally the reference image should cover the area of interest from your input data.
There’s no hard number of tie points that is applied to every data source for every geometric model that I know of. There is a minimum number of tie points required based on the geometric model selection. Discussed in the ‘AutoSync User Guide’.
Ideally the reference area pixel size should be the same or higher resolution than the input image. Too much variance (higher than a factor of 6) between input and reference resolution will have a negative impact on tie point generation.
01-25-2018 07:02 AM
>I have been reading all kind of information about AutoSync what it can do what it cannot do.