I recently did a job using an aircraft equipped with dual 64Mpix cameras. An external L1 gps geotagger was used but ran into some problems. The aircraft flies a line, taking a picture every second or so. At the end of the line, the camera is shut off, the airplane maneuvered to the next line, camera back on, etc. It is a manned UAV, able to map larger areas and we routinely collect several thousand images. Occasionally, the camera's buffers will fill and the front and aft cameras will miss commands to trigger. This results in a rather large mess, as the two cameras will have a different number of pictures. Ultimately this results in a progressively large error in the geotags, especially when reversing lines. We use lots of GCPs...
Our process has been to do an initial alignment using source, find as many GCPs in the photos as possible, uncheck the images, and re-run the alignment. This works pretty well, but occasionally, and entire line will be off on the geotag and will end up floating in space. When mapping non-descript terrain it can be difficult to find good tie point locations.
Is there an better way to figure this out?
Thanks for your help.