The three devices I mentioned each do it differently. A scanning laser radar scans a pulsed laser beam across a field of view in a raster scan pattern sending out short pulses and the receiver records both the reflected intensity value and measures the time of flight for the laser pulse.
A 3D flash laser radar uses a special focal plane array, like a CCD or CMOS camera, except that each pixel has a processing element that measures time of flight of a laser pulse. In this case the laser illuminator is not scanned but has a wide field of view so it illuminates the whole field of view for every pulse. A stereo imaging system consists of two video cameras spaced apart. For each video frame, the two scenes are processed to extract parallax information. Hope that's not too sketchy a description. On May 19, 3:40 pm, Mike Måne <[email protected]> wrote: > > "I'm thinking > > of a 3D imaging system, i.e. one that gives range per pixel " > > How exactly would you do that? Considering an image is perhaps 4,800 > individual pixels, you would have to collect range information for each > 4,800 specks on the camera screen. > > -Mike M > > 2009/5/19 HV <[email protected]> > > > > > > > One of the under-estimated parts of developing weapons systems is > > realistic testing. This is often a source of cost overruns and failed > > defense projects. A system that detects and identifies the priority > > threat in a complex threat-plus-background environment would require > > extensive testing that involves incremental levels of realism, with > > many iterations of test-redesign-test at each level. In this case I > > would think you would need many test sessions involving multiple tanks > > on the move, done at frequent intervals, not once or twice a year. So > > how likely is that to happen? Something to consider, if anyone is > > serious about this. > > > That being said, I think there is an opportunity to develop a sensor > > that could be used for this and many other applications. I'm thinking > > of a 3D imaging system, i.e. one that gives range per pixel as well as > > sufficiently high resolution 2D imagery to discriminate targets. Such > > sensors already exist in the form of scanning laser radars like the > > SICK, 3D flash ladars, and stereo imaging systems. They are > > prohibitively expensive. The challenge would be to reduce the cost to > > an affordable level. > > > Theoretically this type of sensor could provide data to do all the > > required functions: detecting both stationary and moving objects, > > locating them, identifying them, tracking them, and tracking > > paintballs for closed loop fire control. The latter function is > > probably the most stressing for defining the sensor performance > > requirements. It would require a high level of illumination in a short > > pulse to freeze the paintball in flight, otherwise the image is > > blurred, and contrast is too low. It would be much simpler if you > > don't need closed loop control, but who knows? > > -- > -Mike Måne @http://moonrcprojects.googlepages.com --~--~---------~--~----~------------~-------~--~----~ You are currently subscribed to the "R/C Tank Combat" group. To post a message, send email to [email protected] To unsubscribe, send email to [email protected] Visit the group at http://groups.google.com/group/rctankcombat -~----------~----~----~----~------~----~------~--~---
