Medium-term limitations of a large digital system

[horseman 0]

Hi all,

 

I’m a newbie to this forum, but already appreciating it very much.

Already found very good info here, which allowed me to advance with a project I have.

Despite those progresses I’m having some doubts, and would thank a lot all your help, hints and advice.

Please move the thread if not in the appropriate section.

 

My goal is to design and later build a very large surveillance system, to be applied to a school. However, the main objective is not real-time monitoring, even if that can be an equally useful feature to have.

Note: this is a concept, with its execution naturally taking some time, and therefore I’m trying to design it having in mind what is likely to be available in a few years, with technology advancements fortunately making things easier/cheaper by then.

 

Basically, on each of the classrooms of the school would be placed one, two or more cameras, plus a few mics, in a total of around 50 or 60 rooms. So, it would be something between 100 and 150 cameras (by then HDTV), and an equal number of mics.

This is to be installed on a group of buildings in a radius of 100 meters, and all the information should be sent in real time to a central server.

 

The main reason to build this is to control potentially violent behavior from the students.

Example: if an incident is reported by a student or teacher in the classroom X, day X, hour X, an authorized person would be able to access that specific moment in time by grabbing the data from an HDD archive. Having available the various audio and video sources in high quality would allow the correct understanding of situation, and a wise decision. This of course can also work with a robbery situation, etc. It would act mostly as a deterrent system.

 

This kind of concept would needof course very strong and stable software, but that portion is NOT a problem at this point.

The main problem is if I’m correct, the transmission of the data from the high quality cameras to the central server, because as described above, it’s fundamental to send and store a very high quality video file (and audio, but audio is substantially smaller).Once on the server, it could be applied a effective compression codec, without rush.

 

What kind of bandwidth would be necessary, and is it feasible to manage such a high amount of data being constantly sent throughout the network?

And of course, even if not being to build in the immediate future, could the cost be prohibitive because for so much data we could not use “normalâ€

[Thomas 0]

Hi all,

 

I’m a newbie to this forum, but already appreciating it very much.

Already found very good info here, which allowed me to advance with a project I have.

Despite those progresses I’m having some doubts, and would thank a lot all your help, hints and advice.

Please move the thread if not in the appropriate section.

 

My goal is to design and later build a very large surveillance system, to be applied to a school. However, the main objective is not real-time monitoring, even if that can be an equally useful feature to have.

Note: this is a concept, with its execution naturally taking some time, and therefore I’m trying to design it having in mind what is likely to be available in a few years, with technology advancements fortunately making things easier/cheaper by then.

 

The basic building blocks are there. The tech will be cheaper but I wouldn't expect it to be easier.

 

 

Basically, on each of the classrooms of the school would be placed one, two or more cameras, plus a few mics, in a total of around 50 or 60 rooms. So, it would be something between 100 and 150 cameras (by then HDTV), and an equal number of mics.

This is to be installed on a group of buildings in a radius of 100 meters, and all the information should be sent in real time to a central server.

 

I would drop the audio recording requirement. In many areas it's illegal, and frankly it's creepy. The problem here is that you are working in theory but not setting guidelines. Compression methods will be more important then the native resolution of the cameras.

 

The main reason to build this is to control potentially violent behavior from the students.

Example: if an incident is reported by a student or teacher in the classroom X, day X, hour X, an authorized person would be able to access that specific moment in time by grabbing the data from an HDD archive. Having available the various audio and video sources in high quality would allow the correct understanding of situation, and a wise decision. This of course can also work with a robbery situation, etc. It would act mostly as a deterrent system.

 

This isn't future tech. Alot of schools do it now with existing DVRs and cameras. The issue here is more a matter of camera placement then technology. It's not even that rare when you look at real DVRs rather then the cheap ones.

 

This kind of concept would needof course very strong and stable software, but that portion is NOT a problem at this point.

The main problem is if I’m correct, the transmission of the data from the high quality cameras to the central server, because as described above, it’s fundamental to send and store a very high quality video file (and audio, but audio is substantially smaller).Once on the server, it could be applied a effective compression codec, without rush.

 

Or I can compress it at the camera, or I can use a cluster of DVR's with central monitoring software to make it work. My methods will probley be much less bandwith intensive. But I was going to pull raw data across the network, then I would store it in a raw format for later forensics work.

 

What kind of bandwidth would be necessary, and is it feasible to manage such a high amount of data being constantly sent throughout the network?

And of course, even if not being to build in the immediate future, could the cost be prohibitive because for so much data we could not use “normalâ€

[horseman 0]

Thank you very very much, Thomas.

You answered many doubts and made me think about this from another perspective. :up:

 

Regarding the audio capture, I understand your arguments. It raises legal issues and it’s a worry of mine. Agree it can look “creepyâ€

[Thomas 0]

Regarding the audio capture, I understand your arguments. It raises legal issues and it’s a worry of mine. Agree it can look “creepyâ€