I’m not sure what to think about this one. GunDetect is being marketed as a camera that can detect when somebody is carrying a gun. Based on what has been published so far I’m not sure if this is meant to be a legitimate product or a really clever troll.
The first problem regarding GunDetect is technical. Namely what the device isn’t capable of doing:
There’s a question as to how effective this will be as a first line of defense, though. The makers say that their system is accurate “90% of the time” in instances where a gun is clearly visible. That sounds good, but that leaves a lot of room for misses. What happens if nogoodniks are smart enough to conceal their weapons? Also, night vision support isn’t in these existing models — for now, you can forget about spotting thieves in the middle of the night. The technology could easily be useful as an extra layer of gun safety or security, but it won’t replace a good home security system or vigilant parenting.
There’s only 90% chance that the device will successfully detect and gun and then only if the gun is being carried openly and there’s enough light. In other words this device is pretty much worthless at determining whether the person who broke into your home at oh dark thirty is armed or not. But the problems with this product don’t stop there. If you want access to this remarkably limited device you’ll have to spend some major dough. Since it’s 2015 this product has a Kickstarter page. On it you’ll notice two models being offered:
GunDetect comes in two versions, both of which are based on the latest computer-vision algorithms and optical sensing hardware. The difference is the location for the massive amount of number-crunching required to reliably detect a gun in an image.
GunDetect Premium is our main product and does all its vision processing locally using a powerful computing system that does not need to send any video data to the Internet – giving you the peace of mind knowing your private video never leaves the premises.
GunDetect Cloud has less local processing and uses our Internet servers to help crunch encrypted video data – potentially taking longer to detect a gun than GunDetect Premium.
Getting a GunDetect Premium requires throwing $549.00 at the Kickstarter. GunDetect Cloud starts at $349.00 but that only includes a one-year subscription to the service. What a bunch of stingy bastards! The Premium line seems like the only sane way to go since it doesn’t require working Internet service to function, doesn’t upload a constant video feed of your home to a third-party server, and doesn’t involve a yearly $100.00 (I shit you not, the reward tier for an additional year is $100.00) subscription. But for that price you could invest in an actual gun that would at least give you a means of defending yourself against an armed invader.
I don’t think technology able to detect whether is somebody armed is necessarily a bad thing. It could serve as an additional layer of defense for a home or office. However such a device can only be considered effective if it can detect both open and concealed weapons as well as function independently of an external server and not be dependent on environmental factors such as light availability. A weapon detection system that can’t detect conceal weapons is pretty worthless. If somebody is carrying a weapon I can see that already, I don’t need an expensive camera to confirm what my eyes are showing me. Any system that depends on an external server is rendered worthless if the Internet goes out, which can happy for any number of reasons including a burglar cutting your Internet line or the power going out. And what good is a weapon detection system that is unable to detect whether the person who kicked in my door in the middle of the night is armed? That’s the situation where I would most want to know whether somebody is armed or not.
Nothing about this product impresses me. It has technical weaknesses that make it ineffective at detecting weapons, the subscription service for the Cloud model is expensive, the price of the standalone Premium model is very expensive, and the Cloud model creates some serious privacy concerns. Judging by the number of backers so far I’m not the only one who sees this product as a nonstarter. If this is meant to be a legitimate product it would behoove the developers to return to the drawing board and sort these problems out before begging the Internet for money. If this is meant to be a clever troll I must tip my hat to them.
A high quality IR camera coupled with this software might be able to do a decent job but the sensors are well above the price point.
Even cheap cameras generally pick up infrared unless you apply a filter, which means adding infrared LEDs to enable night vision would add a significant amount to the cost. But it seems to me that this project is about pumping people for money more than doing anything useful.
Oh I know most CCDs are infared sensitive its just the level of sensitivity to pick up a gun through clothing means a high quality IR specific sensor which is prohibitively expensive, and it would not require an IR illumination source.