Home
Video Security: Network Cameras; Network Recorders; Monitoring Software; Network Encoders; CCTV Cameras; Accessories; Warranty and Support Agreements

The DEPA approach to video security

Discover a smarter approach to intelligent video analysis with Sony’s Distributed Enhanced Processing Architecture (DEPA). Find out just how it works, the benefits for your business and how one company is providing greater security to its customers as a result.

More intelligent intelligence

Video security was once very dependent on the human eye, with someone sat in front of a bank of monitors on the lookout for threats or incidents. More recently, the introduction of intelligent video analysis has helped reduce the burden on security staff through the triggering of pre-set alarms.

However, the cost of intelligent video analysis can be prohibitive – especially in an analogue system – as it typically requires expensive back-end processors.

Sony takes a different approach. Our intelligent video analysis is built on the Distributed Enhanced Processing Architecture (DEPA), which distributes image processing across the network on multiple low-cost processors, resulting in a security solution that’s more flexible, scalable and economical.

SNC-DM160P

Spreading the load

Video analysis comprises two stages: the pre-processing stage, to analyse images and generate metadata, and the post-processing stage, to store, filter and raise appropriate alarms. In an analogue system, these are both performed on centralised processing units at the back-end of the system, which often creates network bottlenecks and inefficiencies.

In Sony’s DEPA system, the two phases are physically separated and performed on different system components. The pre-processing is performed at the front-end on the cameras themselves, and the post-processing is performed at the back-end on low-cost DVRs.

Front-end pre-processing

Sony’s DEPA-enabled security cameras use Intelligent Motion Detection (IMD) and Intelligent Object Detection (IOD) to pre-process the video stream.

IMD is a motion-detection technology that’s unique because it can distinguish between “real” movement and “environmental” movement, such as a tree blowing in the wind or ripples on the surface of a lake. This minimises false alarms and operators spend less time searching through video to find real events.

IOD can detect any object that enters or exits a camera’s field of view, and then alerts security staff if it breaks a pre-determined rule – for example, if a car is left in a no-parking zone for a certain length of time or if an item in a shop is picked up but not placed back.

If a potential threat is detected with IMD or IOD, it is transferred to the back-end system as metadata for post-processing.

Back-end post-processing

The post-processing phase in a DEPA system is performed on a DVR running Sony’s RealShot Manager software. It involves filtering video for potential threats and raising alarms if a pre-set threshold is breached, for example, if an object appears in a designated area or if an object passes a virtual line.