Industrial CCTV Surveillance System using smart video analytics to safeguard against fire-theft-intrusion and sabotage.

Protection of Factories and Industrial Assets

• Intrusion detection
  
• RFID and CCTV Integration
  
• ANPR
  
• Fire & Smoke detection
  
• Stock Count
  
• Yard Security
  
• 4G Live Monitoring & Alerts
  

The basic definition of video analytics, also known as video content analysis or VCA, is the extraction of meaningful and relevant information from digital video. 1 It is the capability of automatically analyzing video to detect objects, characterize their appearance and movement, identify individual objects, and recognize their patterns of activity. Video analytics technology has broad application beyond the security industry, including in the following industries: 3-D gaming, animal behavior research, automotive and railway transportation, entertainment, health care, horse racing, industrial safety, manufacturing, mobile policing, retail, robotics, sports, traffic enforcement, and wildlife management. The broad range of applications provides good incentive for global investment in video analytics research and development (R&D), many aspects of which will benefit security video analytics.
Advances in video analytics technology are made possible by advances in the underlying information technologies, such as computer processing power, memory speed, parallel processing, solid state and hard drive data storage, high-speed databases, high-capacity networking, and artificial intelligence (AI). Video analytics is actually a subfield of AI. Video analytics capabilities fit both of the definitions of AI below.

he capacity of a computer to perform operations analogous to learning and decision making in humans, as by … a program for the perception and recognition of shapes in computer vision systems.2
The ability of a computer or other machine to perform actions thought to require intelligence.3
The ever-accelerating pace of advancement for the underlying technologies provides the research and development (R&D) efforts in video analytics with ever more powerful tools with which to develop and refine video analytics applications.
Two important video analytics improvement trends include higher detection rates and lower false alarm rates. In the U.S. National Institute of Standards and Technology’s (NIST) assessment of biometric face recognition in still images, the error rate halves every two years. In 2010, the best face recognition method matched 92 percent of mug shots to one out of 1.6 million images.4 Since then, the video analytics technologies have continued to advance in quality and effectiveness.


One of the largest guarding companies in the world is already using video analytics to support its operators and as a backup for patrolling guards. Some smaller, specialized companies are even using the technology as the foundation for their entire business of providing security and remote guarding as a service.
With the help of strategically placed network video cameras and advanced video analytics, fewer operators can monitor more and larger installations 24/7, and they can react more quickly and adequately to incidents. Automatic alerts flag suspicious events in real time, including a person moving in or near restricted areas, loitering, or someone trying to tamper with property, including the video cameras themselves.
Once the video analytics application has alerted an operator about an incident, the operator can verify the alarm before sending a security guard. This way the number of false alarms and unnecessary emergency responses is kept down, and the guards can focus on the areas where they are needed most.
Sometimes, the only response to an incident that is required is to deter the intruder by letting him know he is being watched and recorded on camera. Connected horn speakers allow the operator to address the intruder, while the cameras transmit the intruder’s reaction. Not having to send out a guard in these instances can saves hundreds of dollars – a saving that is ultimately passed on to the end customer.
Video analytics applications are developing fast, and they are changing the way the security industry works. They are not going to replace security guards entirely, though – ultimately, it is always up to the operator to validate any alarms raised, and make the right decision on how to react. 
In 2013, the world’s largest supplier of automotive components (which also provides 90 percent of Google’s autonomous driving software) announced that it had begun sharing video analytics technology with one of its sister companies, a security systems manufacturing company. This is a noteworthy example of how the security industry will continue to benefit from analytics advances in other industries, some of which — like the automotive industry — have R&D initiatives that are significantly better-funded and better-staffed than R&D initiatives within the security industry.
Overall, four key factors are driving the ever-accelerating advancement of security video analytics:

• Information Technology Advancement. Across-the-board advancements in the information technologies with which video analytics offerings are built.​​
• R&D Investment. For many industries, there is growing investment in video analytics R&D initiatives, with hundreds of notable universities and research institutes around the globe participating in the efforts.​​
• Increasing Risks. The rising levels of homeland security and critical infrastructure risk — and the size of the physical landscapes that require effective video surveillance — heighten the need for more effective and more capable video analytics.
• High Security Industry Adoption Rate. Video analytics features are becoming commonplace in 
• consumer- and commercial-grade security video products, and are demonstrating high value compared to earlier generations of video analytics technology. ​
• These are among the many reasons to gain a clear understanding of where video analytics technology is today, where it will be tomorrow, and how to evaluate analytics companies and their technologies so as to select the capabilities that best fit a specific organization’s security and business needs. Today, there are many more considerations involved in video analytics selection than with the previous generation technology, due to the nature of video analytics evolution.
  
  
  How Video Analytics Has EvolvedIt is worth taking the time to understand the evolution of security video analytics, which requires examining the following aspects of the technology:
  
• Stages of security video analytics architecture
• Metadata advancement
• Accelerating trend in error rate reduction
• Stages of Security Video Analytics ArchitectureVideo surveillance systems started out as fully analog systems known as Closed Circuit TV (CCTV) systems. In these early systems, one camera connected to one video monitor. There was no recording. Gradually, surveillance systems evolved into the fully digital, computer-based, networked systems that we know today — systems that are based on advancing information technology.
  Video analytics functionality is provided by software. Software runs on a computer or a device with computer circuitry in it. Thus, security video analytics software runs on the computers and devices of a video surveillance system. As the computer and networking technologies of surveillance systems evolve, so does the design of video analytics software. However, the design of security video analytics software does not evolve just to keep up with changes in computers and devices, it evolves because of the need to improve surveillance capabilities. To date, video security analytics have been evolving to take maximum advantage of the computing capabilities of security computers and devices. However, the high value of video analytics is now beginning to influence the design of devices that run security video analytics, including cameras.
  
  
  
  Physical ArchitecturesIt is helpful to view security video analytics advances as having evolved into five physical architectures. The word architecture means “the carefully designed structure of something.”5 This paper uses the term physical architecture to refer to the physical device or computer where the analytics processing takes place. The first four architectures listed below, began with the previous generation of security video analytics, where the analytics processing was performed completely within a single device or computer. The hybrid architecture is a recent arrival, and refers to the system design in which some parts of the analytics processing are performed on one device, and other parts are performed on one or more different devices. This is a significant technological advancement, and will be examined in detail.
  One physical architecture does not necessarily replace another, and all architectures have continued evolving to a greater or lesser degree over time. The order in which they are listed below is the order in which the architectures have come into use. They are:
  
• Appliance-based
• Server-based
• Camera-based
• Cloud-based
• Hybrid