Ok, this calls for some innovation... Unless its been done effectively before, you could come up with a patent ;-)

The most long-life solution would be those where the sensing devices are not destroyed by the bullets, such as a video camera or some other source of image processing. If you could bulletproof it, the array of LED-sensors is the best option actually... considering the caliber of the bullets, you wouldn't need a large number of emitter-detector pairs. Since this option is already used for several touch screens, you can actually use a commercially available solution. However the speed through which the bullet passes through could be major issue. Your microcontroller or sensing device should be able to catch the extremely short transient flicker in the signal.

However if your company wants to keep milking the client for a long looooooooong time, I'd say the best option would be a destructive sort of method where only you can keep providing new targets when the old ones wear out :-))

Here are some other RADICAL ideas for non-destructive type target capture:

* Sound Processing - would it be possible to deduce the position of the bullet using an array of microphones? Do whizzing bullets produce enough sound? We dont hear it, but maybe thats because its out of human hearing range? At least two sensors would be needed. The sound signal will be filtered out to the range made by whizzing bullets, and the relevant position of the bullet would be when the signal is at a maximum. This would involve *MAJOR* DSP and high end procesors, and you would still have to bulletproof the microphones, but then the better to milk you client with ;-)

* Radar :-)) - related to sound processing, only there will be a radar transmitter and receiver.

* Mount a device on the rifle barrel itself. This will be like those targetting lasers, only the shooter should not be able to see where they point. This device would also involve some image processing though.

These alternatives will have a very high intitial cost, but then they will be cheaper in the long run, since they client doesnt have to keep replacing electronic targets.

The other alternatives would be, as Steve suggested, where the target itself provides the x-y signal, but gets destroyed at every shot. What if the target is a bulletproof sheet of metal with a matrix of strain gauges on the other side? The strain gauge will doubtless sense the smallest bumps on the target, but bigger bumps will be a problem.

Other things you have to consider is that if the targets will mostly be out in the open, they have to withstand all sorts of atmospheric conditions.

Do let us know what you finally decide to use (if you are allowed to!), we'd be most interested to know ;)

kundi