Despite of having fully implemented CCTV Cameras, many major and provincial cities are still faced with lack of security and safety. What are the future research directions in securing unsafe neighborhood with CCTV Cameras?
Not my main area, but noticed your question and had the following thought: you can cover an area with CCTV cameras (the 3-D coverage problem is known in the literature) AND you can feed all these camera feeds to a central console. The missing link for me is the need to have intelligent interpretation of all that data! Perhaps face recognition software can help? Also, if no one is watching the video feed there is no security. You might as well mount dummy cameras on the exterior of a few buildings.
(You could have a strategy of putting up many cameras, and selectively monitoring a sample of them; the bad guys would not know which cameras to avoid?)
Most modern cameras are much more than cameras. They include dry contacts to turn on / off almost anything, the ability to record and play sounds, sense temperature, receive digital and analog inputs and communicate with a variety of other devices.
The back end management platforms, like Occularis ( as just one example ) and larger scale systems, can perform event driven programming, and a variety of interactive operations, It is much more that can be done that using these systems as a video camera recording mindlessly while someone ignored hundreds of screens.
That the cameras themselves have local computing power, some with built in facial recognition and object tracking - they can be programmed to accomplish significant things in the absence of the control / management plane.
I have seen off the shelf, commodity cameras capable of detecting a pedestrian trying to enter a walkway against traffic, that event signaling the traffic signal to enter a all-ways-red state, and visual and auditory alarms being sounded via dry contact closure to keep the pedestrian from being run down. This is not some future technology - its something that can be done today. This happens in real time, without the management platform being involved. It is a highly survivable system, and not dependant on some far flung data center or cloud. It would work in a remote village or a busy urban population center.
( crappy ) PTZ cameras are giving way to multi-lens 360 degree cameras. So instead of monitoring one place or a series of places along a tour, the entire area is being monitored - even places not anticipated by those designing or deploying the system. These systems are providing much better resolution too. So what had to be accomplished with zoom lenses can be done digitally - to any portion of the scene - after the fact.
There is the concept of "blank screen monitoring" - So imagine a person monitoring the cameras. Its not a millions cameras on a millions screens as seen in The Matric. Normally their screen is blank - completely blank.
They can choose to look at a given camera or set of cameras, but normally, they are not having to hop through all the cameras looking for bad stuff.
The CAMERA ( or the management plane ) detects when something of interest is happening. This may be something where it shouldn’t be, motion when motion normally does not occur, someone pressing a butting, making a sounds, fire, door being opened or closed, presenting an ID card, being detected by a wireless network, ...
These anomalies trigger the management system to display THAT camera. It can also force the operator to acknowledge the event. It can correlate those events over multiple cameras. As additional cameras require attention, they are tiled and then spill over on to other screens.
In one application, the operator has the ability to send screens to a mobile app of an officer or partol car. So now the officer on site has the same view of things as the central office. This is incredibly powerful. Its also easy - no longer does the dispatcher have to try and narrate what is happening - which keep sthem from paying attention to other situations in progress,
There is an enormous amount of things which can be done today with shipping cameras and code. The problem is that most safety offices view them as dumb cameras. They do not have someone monitoring them. they have not seen a better way to do it. Also many of the places which do this well - casinos most notably - are not really willing to share their experiences in the interest of keeping their edge over bad actors.
Part of the reason nobody knows about this is that rarely do you find safety offices that (A) value this sort of interactivity (B) have a dispatch staff capable of making it make sense.
https://scholarcommons.scu.edu/cgi/viewcontent.cgi?article=1005&context=idp_senior
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India