sexydadee 0 Posted June 24, 2012 I am wondering how intelligence at the edge works. If the camera is doing all the processing, where does it keep the intelligence gathered? does it send it along with the video signal as a seperate type of data or hardcoded with the video? How does the "other end" (ie nvr, client operator) interpret the intelligence? is there a standard or only the "other end" of the same manufacturer of the camera should be used? does the nvr recognize that it is receiving video+intelligence? I'm confused. Appreciate any feedbacks. Share this post Link to post Share on other sites
Soundy 1 Posted June 24, 2012 How are you defining "intelligence"? "Intelligence at the edge", AFAIK, is just a buzzterm that generically means that the camera is doing all the motion/analytics/etc. processing; it doesn't necessarily define what it does or should do with the processed information afterward... which means the answer to all the rest of your questions is, "it depends". Share this post Link to post Share on other sites
sexydadee 0 Posted June 25, 2012 ok, let us say, motion detection. what does it do with the motion it detected? Share this post Link to post Share on other sites
Soundy 1 Posted June 25, 2012 Well again, that depends on the camera and NVR/VMS. Most commonly, it's used to trigger recording, or to start sending a stream to storage. It may also be used to send email/SMS alerts, or to trigger an alarm output, or any number of other things that the camera's manufacturer may program it to do. Share this post Link to post Share on other sites
akelley 0 Posted June 26, 2012 If you're referring to simple events such as motion detection (MD) or detecting external triggers (alarms) an NVR is simply notified (via the connection stream) that an event has occurred, which the NVR then determines what to do, based on that specific event. For example, a MD event could mean that the NVR is to begin recording the video stream at 30 fps, send an email, and send an X10 command to turn on an exterior flood light. So the camera just provides the event trigger based on the camera processing the MD event, and the NVR decides how to respond to it. You may have read some threads on this board where people ask which NVRs are capable of detecting motion on the NVR or having to rely on the camera's on board processing to detect motion. There are benefits to both, of course. Other types of analytic you may be referring to such as people/object counters, people/object trackers, etc. are manufacturer/camera specific. For example, Axis has a development platform that you (or third parties) can write these types of applications that run on the camera. Based on your requirements, these cameras can simply send out event triggers or send specific data (i.e., the number of people counted going through a door each hour) to an NVR or other computer that would know how and what to do with this data. You typically only find this on-camera processing in higher end cameras. Intelligence at the Edge simply refers to the camera being able to do this type of processing. Storage at the Edge refers to the camera being able to record/store video on the camera (SD card) or to an attached storage device such as USB drive, NAS, etc. Share this post Link to post Share on other sites
sexydadee 0 Posted June 27, 2012 I see, so i guess youll need some special software or NVR to be able to interpret whatever it is the camera is sending aside from the video. thanks all for your replies. I am more enlightened now regarding this matter. Share this post Link to post Share on other sites
