Given the wide variety of scopes, tracking abilities, and CCD imagers out there, it seems like your plan would need a minimum requirements package for those wishing to participate, some fairly good training, and coordination of which areas of the sky are being covered. I suspect there aren't too many of us around with the kinds of imaging skills that would produce worthwhile, consistent image data as a first hurdle to consider. Actually, from what I have seen so far, the quality of scopes and the CCD technology is already at an acceptable level for an automated process...This whole system would only be intended to locate possible NEOs...Any promising results would obviously have to be verified by the big boys toys...
My intention of having this thing fully automated would require a GoTo Scope...But, anyone cable of capturing a CCD should be able to submit their images...The system would just need to know the general sky coordinates and the time that it was caputered for it to be considered a useful image. It would be my desire to see this project developed with open technology so that the companies that are already developing software for scopes could build this API (Application Program Interface) right into their already developed software...All they need to get from the system would be a potential coordinate...Then they simply submit that image back to the system with a time index of when it was captured...
The verification of received images sounds like it would be a fairly tedious and time consuming task. And the automation you describe for feedback and control would require something fairly sophisticated and consequently costly. You're not thinking of having everyone running a Paramount connected to the web are you? Or does your proposal perhaps include upgrades for commercial GPS goto scopes? The verification process really doesn't concern me too much. I have already done similar things with my 3D processing tools. I would only have to look at the predominent stars/galixies and match them up to the master image...Once the predominant characteristics of the image are verified and matched with the master image, I could factor in what ever scale/rotation factors needed to match the remaining portions of the image...In other words, the verification tool would be impervious to different rotations/resolutions/scales of the submitted images.
At this point, the system would only look for anomylous objects that can't be verified in the master image. These objects would be categorized as anomylies. In order to properly determine if this was in fact a NEO, I would take all the anomylies and calculate them over time...At this point, it is simply a process of elimination... <<<WARNING, Programmer language comming>>> Each anomyly would be linked to every other anomyly taking into consideration the time that might have passed from one submited image to the next. Assuming that every anomyly could be a NEO, the system would just keep linking every new anomyly found to every existing anomyly previously found. In order to start eliminating these links, the system only has to do some simple 2D math calculations taking onto consideration of when the submited picture was captured...Anything that doesn't demonstrate a smooth linear trajectory can quickly be unlinked...Over the course of several matched images, you would quickly be able to determin the NEOs from anomylies...This system would be able to withstand the submission of numerous poor quality images becasue it is the overall number of linear links on each anomyly that increases it potential of being a NEO.
Lastly, it seems to me you don't want to include too many details in your proposal to NASA, or someone there may run with your idea in-house and never bother to tell you if it had merit or not. Actually, I don't care if someone runs with it or not...I think the important thing is that we do whatever possible to detect the NEO. However, I would sure love to be a part of it. Given what I have already done in the past, this would be no different than what I have already done several times over...