Today, the video feeds from millions of surveillance cameras are just recorded to facilitate post-event forensic analysis. Such a passive system cannot help with crime or terror prevention, accident and mitigation which require real-time event recognition. I4S India Editor gathers insights from Giri Krishna, CEO, Silvan Innovation Labs, where he tries to define a roadmap for products that need to be developed that will address the specific needs of the Indian market and situation on the ground. Excerpts:
Let’s begin with an overview of the IP based surveillance systems
An urgent need has been felt to move away from traditional systems and towards digital IP based surveillance systems as they are expected to provide the ability to respond to events as they occur and also enable remote monitoring even as the affected areas remain inaccessible. This migration is not without its challenges.
Typically, CCTV systems are deployed over a number of remote locations and are monitored by security personnel at a central location. In such systems images from multiple cameras are streamed to a video monitor that can be viewed by a security person. As the number of cameras increase, the potential for the security personnel to miss a significant event increases. In addition video recorders are used to store video images from the cameras for future analysis. These video recorders have capability to analyse the videos, detect events and trigger alarms. This reduces the potential for mistakes by the security personnel.
In IP based surveillance systems, the effectiveness of such a system is determined by the availability of the network as well as the bandwidth available with the network. The fact that the video is streamed over the network can result in delayed response times to events that require immediate action.
What do you think are the challenges of an IP-based surveillance system?
The number of failure points in such a system can also lead to an ineffective surveillance system in conditions where bandwidth is limited and power availability is an issue. There is a possibility that critical video information is lost and important events are missed whenever there is network or power failure, in a centralized surveillance system.
The key problems in deploying an effective IP based video surveillance system in a developing market like India is the availability of 24/7 power as well as network bandwidth. For the surveillance system to be really useful it is important that critical video information is not lost due to non-availability of power or due to limited network bandwidth. Networks are also known to go down on occasions and hence the surveillance system needs to be designed to be functional in all such situations.
It would be desirable to provide a distributed video surveillance system that has cameras with the ability to address situations involving lack of power and limited or no network. In particular in the case of power failure, cameras need to have a local independent power source that will ensure that the camera continues to function. In the case of network failure, the camera should have the ability to store the video images locally and the video content can be synchronized with the central location on resumption of normal services. In the case of network bandwidth being a limitation, the camera should have the ability to decide the resolutions and the frame rate for streaming the video over the network.
To improve response time to specific triggers or events, the camera should have the ability to analyse the video, detect specific events and trigger alarms. The same mechanism can be used even in situations wherein network failure has occurred. This system ensures that the camera is the only point of failure and the other cameras in the system can takeover if one camera fails. As a result such a surveillance system becomes highly available and reliable.
Cameras will need to have both battery backup as well as local storage to provide for an eventuality where either power and/or network are unavailable or limited. To support the effective utilization of the battery backup and local storage, algorithms will need to be implemented to support multiple situations. While providing battery backup on the camera is an obvious solution for enabling cameras to continue to work in the absence of AC power, power utilization and optimization will be required to extend the charge on the battery so that the cameras continue to function even if power remains unavailable for extended periods. Some of the features that need to be considered are:
– Sleep mode with ultra-low power consumption: When no motion is sensed, the camera can be put into a deep sleep mode to bring down the power consumption drastically. This helps in increasing the life of the battery.
– Quick wake up: The camera needs to be equipped with motion sensing capabilities that enable it to detect the presence of an intruder, bring it out of sleep in milliseconds and take the required images or video.
When there is a power failure and/or if the network is unavailable, the system should be able to mark the point of failure to resume transmission over the network when it becomes available again. In situations where bandwidth is limited, the system should also be able to make decisions with respect to the frame rates and resolutions to be used for video streaming.
Whenever a specific event of interest occurs, the system can be programmed to switch higher frame rate and/or resolution to enable a clearer view of the video at the viewing station. The usage of intelligence and algorithms to optimize utilization of power and network by the surveillance system has been term Operational Analytics and tremendous opportunity exists for developing products with operational analytics for the Indian market.
How does video analytics address infrastructure challenges?
Video analytics addresses infrastructure challenges by enabling content-aware storage and routing. The software is the equivalent of having a tireless, unblinking person watch each camera, detect relevant events, and make decisions on what events to store and when to stream video.
An analytics-enabled Video Management Software (VMS) addresses the three big scalability challenges of enterprise video deployments – human resources, storage and bandwidth – more effectively and cost-efficiently than a system without this capability. In view of its importance, it is an urgent necessity to come up with video analytics that can autonomously track, monitor, analyse and verify real-life situations in order to alert in the event of any abnormal activity without any delay. As the importance of autonomous large scale surveillance system continues to rise, especially in homeland security, video analytics is bound to take a pivotal role. The success of autonomous surveillance system solely depends on video analytics, which is the brain behind cameras.
Embedded video analytics provides a win-win situation, by enabling security personnel to prevent crime instead of investigating it after the incident. A distributed video surveillance solution greatly reduces operational cost while increasing the availability, reliability, scalabaility and performance of the system while allowing centralized administration and monitoring. This technology can transform the way enterprises plan and deploy IP video surveillance systems. Systems with these features can be deployed in public places like Airports, Railway Stations and Bus Stations as well as large campuses including places like IT Parks to prevent as well as limit the impact of terrorist attacks.
Video surveillance on the digital IP-based network environment is highly dependent on availability of network bandwidth. This constraint needs to be addressed by relocating the intelligence from the remote analytics center to embedded intelligence in the camera thus providing alerts in real time with high levels of functionality with significantly lower costs and bandwidth requirements. The device at the edge should have analytics that can detect events in real-time and stream video over the network for secondary processing on the server side. The video management software (VMS) in the system will have the intelligence to make decisions with respect to usage of storage and generation of alerts and alarms.
How to mitigate the challenge of shortage of technically skilled staff?
One of the key barriers to deploying IP based surveillance systems is the need for a higher level of technically skilled support staff. Since creating a pool of trained manpower is going to be a challenge, systems need to be designed in a manner such that installations can become a trivial task. IP based systems should require zero-configuration and as the move to wireless IP picks up devices should have the ability for self-discovery and self-configuration in the network.
There are specific challenges that need to be addressed to enable a seamless and smooth transition from analog systems to digital IP based surveillance systems in the Indian market.
What are your views on IP-based surveillance system in India?
Various reports suggest that the total India surveillance market is around Rs 2500 crores, growing at the rate of 25-30% per year. Of that the IP camera market is Rs 300 crores and the CCTV camera market is Rs 1500 crores. The market share of IP-camera market is steadily increasing, and by 2012, IP-camera volume is projected to take over CCTV (analog) camera volume.
The Indian market is mostly served by imported cameras. The video analytics has not yet picked up the demand in commercial surveillance market, but there are specific needs in Defence and Traffic management. The market is currently catered by numerous small service providers/dealers who do the system integration and installations of mostly imported cameras.
Recent security issues, government regulations, and starting of ‘surveillance as service’ business by few service providers will create a bigger discontinuity in the market. Clearly an opportunity exists to create products for the Indian market that meets this need but also to grow the market at a rate that is much faster than projected.
