Shankar Viswanathan wrote: > One thing I have heard about using webcams as security cams is > reliability. Consumer grade webcams aren't generally designed for 24/7 > operation...
That's sad if a device with no software and no high wattage CPU can't hold up to 24/7 operation. I have a couple of bottom-end TRENDnet IP cameras that tended to overheat and crash, but they had a CPU and ran Linux. (These cameras were dirt cheap and still not worth what I paid for them. I missed the retailer's return window, and they've since bricked themselves. Thus I refer to them in the past-tense.) > He uses the 'motion' package to do motion detection... > He's quite happy with the results, with two caveats: > Any large vehicles passing by on the street during the daytime also > triggers the motion detect. 'motion' is supposed to be a "cutting edge" solution for this, and he has it running on a device that outperforms most IP cameras, which seems to suggest that software-based motion detection just isn't practical yet. (Though perhaps there are expensive commercial packages that do a better job.) I had no end of problems with with software-based motion detection built-in to my IP cameras, and ended up attaching a passive IR motion detector to the GPIO input port on my camera that supports that. This has worked perfectly with zero false alarms. Makes you wonder why IP camera vendors can't be bothered to spend $2 or $3 for a built-in PIR sensor. Panasonic does it. Ben Eisenbraun wrote: > Curtains. I struggled with the curtain solution too. Using something light diffusing, but not light blocking. It didn't resolve the false motion triggers, but did compensate for the poor automatic white balance, which used to cause the camera to completely wash out when there was full sun coming through the window. -Tom _______________________________________________ Hardwarehacking mailing list [email protected] http://lists.blu.org/mailman/listinfo/hardwarehacking
