Currently, orthorectification in ENVI is only done through RPCs but any sort of orthorectification is going to require knowledge of the sensor or camera. I have not seen the warning message regarding the information not being accurate enough, so I suspect what it says is true. Usually, RMS is quite small when you save the RPCs to the file in the final dialog. In the GCP dialog, an error of 1 or less (depending on the resolution of the data) is desirable. I am wondering if it might be the camera information that is insufficient.
|