IT Enablement in Video Surveillance
Early forms of video surveillance were mainly used in industrial management. With feeds sent to a centralized control room from video cameras placed at different locations, irregularities could be discovered and promptly addressed. This reduced major accidents while providing a better way to identify causalities and avoid future problems. Video surveillance equipment became especially useful in difficult-to-access areas or spots that personnel could not continuously monitor. With the ability to provide evidence before and after incidents, video surveillance became widely applied in public security and other fields, eventually becoming known generically as closed circuit TV (CCTV).
Early in the 21st century, many localities began promoting Safe City projects with video surveillance at their core. High-ranking officials stress that IT is a main feature of the technological revolution bringing profound change to work, social, and family life. Many of these same officials say IT injects new vitality into Safe City efforts. To this end, law enforcement and other governmental departments are integrating advanced technology and information support into Safe City projects. Smart cities rely on these platforms to ensure public safety, provide evidence in court cases, and manage traffic.
Smart alarms and surveillance systems integrate security, computer applications, network communications, video transmission, and access control into one platform. In addition, video digitization, network-based applications, and advances in technology are becoming more sophisticated and extensively deployed. Uses now include areas such as public security, transportation, and a long list of other industries and sectors. These systems combine cloud computing, the Internet of Things, Big Data, and other forms of new-generation IT to enhance cross-departmental data sharing and improve the efficiency of municipal operations.
IT and video surveillance emergence
Video surveillance technology is a combination of various elements. When video and monitoring technologies first appeared, CRT monitors were the main vehicle for viewing, coaxial cable was the main medium for transmissions, and tape was the major storage media.
Those days are gone. IT-enabled video surveillance is a natural byproduct of the new technological revolution. As technology has become more digitalized, network-based high-definition has expanded into security system applications. Connectivity is also becoming a more important part of the deep dive into IT enablement.
Surveillance videos and images contain huge amounts of information. They are prime candidates for the massive information processing focus of IT. As video surveillance comes into wider use, surveillance systems will transition to digitization, network-based operation, high definition, high integration, and intelligent functionality. Sharing video assets and supporting multiple services will also be on the agenda.
HD video transmission, forwarding, storage, and supporting software and hardware pose intense challenges. As a result, system stability, reliability, and application practicality are a concern. Conventional IT architectures, let alone primitive CCTV monitoring systems, cannot meet the large-scale deployment needs of HD video surveillance applications. Internet technology, cloud storage and computing, Big Data, and other cutting-edge technologies are being applied on much larger scales, making IT enablement for video surveillance more evident.
Trends in IT-enabled video surveillance
Trends in IT enablement for video surveillance are most noticeable in 4 basic ways.
Analog to digital transition
The transition from analog to digital, and then to IP networking, has made IT technologies ─ cloud computing and storage, Big Data, video analysis, and intelligent surveillance technologies in particular ─ the main engine behind advances in video surveillance technology
Many large IT companies are entering the video surveillance industry and cooperating with security companies in multi-faceted areas
A growing number of technology-based security companies are hiring large numbers of IT professionals and focusing on digital signal processing, cloud storage, video analysis algorithms, Big Data, and other core technologies
IT enablement relies heavily on Internet connectivity to support the required long-distance transmission of video images. Most mainstream network monitoring systems require a powerful platform for control and scheduling management and the software and hardware infrastructure adopts an extensive amount of IT technology. In addition, other factors are having a profound effect on the development of video surveillance technologies, including the:
- Move to HD
- High integration and intelligent functionality
- Massive amount of information that must be transmitted and stored
- Application of video analysis technology
- Deployment of cloud storage, cloud computing, and Big Data
Video cloud is creating a storm
Cloud storage and computing are based on distributed networks and provide storage and computing resources as a service. Customers store data and execute their computing operations in the cloud, encapsulating the "network is more like a computer" concept. Being more energy efficient and reliable, cloud storage technology is set to become a fundamental part of Safe Cities and other surveillance efforts.
Analysis of surveillance video entails a large amount of data that must be processed at high speed. Consider the vast number of features and individuals that must be compared in facial recognition applications. Even if multiple high-performance servers have a conventional parallel processing scheme, the computation speed is still not fast enough to satisfy the time requirements for some customers. Cloud computing provides the solution in such instances to deliver vastly enhanced efficiency in the calculations.
The cloud storage and computing model will continue well into the future for data-intensive video surveillance. However, one concern is that data processed and stored on a remote cloud lacks the security of a physical boundary, which means that subscriber data management and ownership are separated. This model is more prone to data leaks and tampering. Network security and privacy issues have become a major hindrance to the continued development of cloud storage and computing and these issues will continue to be a major point of attention until a fully viable solution is available.
Big Data refers to the copious amounts of data involved in an analysis that makes review by a human or group of analysts lack any sort of efficiency in terms of intercepting, handling, collating, or processing the information.
Big Data technology can produce sharp advances though. Four main themes emerge.
- Intelligent analysis: Intelligent analysis uses computer-based image recognition algorithms to analyze video content and generate a structured or semi-structured text-based description of content and behavior for Big Data analytics.
- Distributed processing: This technology is based on large parallel or distributed processing technologies such as Hadoop or Hbase. They form large distributed data storage and management frameworks that provide real-time and batch Extract, Transform, and Loading (ETL) plus data scrubbing, conversion, and load balancing.
- Data mining and analysis: Data mining and analysis is the highly precise use of converted video metadata. Video data is converted to text and used to correlate the events and patterns in content.
- Visual representation (visualization): Visualization technology helps efficiently search, compare, and display video content. It significantly enhances how images and data are presented in information analysis, emergency response, Geographic Information Systems (GIS), and other systems.
Most of the data generated in video surveillance is unstructured and the volumes of content are usually loosely coupled, challenging traditional approaches and mechanisms in data management. The whole basis of Big Data-based architecture relies on putting vast amounts of data into smaller, easier to access batches, then processing it on multiple servers with parallel analysis capabilities. This model provides a huge advantage in turnaround and quantitative analysis over humans.
Big Data can be summarized with the 4 "V"s:
- Volume of data
- Variety of coding and data structures
- Velocity of intense processing
- Value of extracting useful information
In video surveillance, the value of data is often inversely proportional to the size of the volume and density of the data. In one hour of video, there may only be a few seconds of useful data. Using powerful algorithms to accelerate clean up and filtering, then extracting the value content is something that Big Data tries to solve. It is paramount this analysis be completed with efficiency and accuracy.
Higher-density information is often the most valuable information. Mathematical and statistical theory points to this phenomenon. In terms of video footage, a higher-density clip with movement would likely be more pertinent than a static segment without any events or movement. Real-time digesting of huge amounts of surveillance data produces large amounts of dormant or static data that wastes a considerable amount of storage.
At the same time, accuracy and efficiency of the video analysis determines the value that can be extracted from the data. Customers involved with Safe Cities commonly require low latency and enhanced accuracy. The video surveillance field is in urgent need of Big Data technologies that can intelligently analyze and extract value from voluminous frames and segments to provide summarized information and reduce the amount of data human analysts must process. To this end, analytics must provide a metadata repository to improve the efficiency of using the information.
The move to IT in video surveillance
The move to IT enablement in video surveillance is demonstrated in improved functionality and feature-rich applications. After Safe Cities become fully digitalized, network-based, and intelligence-rich, the overall layout will become more expansive in coverage. It will feature hyper-connectivity and become highly accurate and efficient with smart capabilities.
The application layer will have powerful Big Data processing capabilities to digest and collate the massive amounts of information found in video feeds and archives. These capabilities will help enable preemptive warnings and early intervention.
Blending Big Data and video surveillance technology with cloud computing and storage technology that is specifically geared toward video and image processing can achieve the purposes of the Safe City agenda. Cloud computing, cloud storage, and Big Data are currently the fastest growing areas in IT technology. These technologies are also tools that must be more vigorously applied for further development.
IT-based developments in video surveillance technology are sure to add momentum to enterprises involved in home security and Safe City programs; especially as monitoring and IT companies cooperate more closely. IT companies are more apt to make product upgrades as technology changes than security companies with little IT know-how are. At the same time, IT companies may not have the necessary knowledge to engage in the security industry. For this reason, it is as problematic to depend solely on IT enterprises to apply Big Data applications, as it would be for security vendors to attempt system integration without the help of proficient IT counterparts.
In the Big Data era, almost every field is becoming interrelated with dozens of others. Video surveillance is no exception. Video surveillance must continue technological and industry integration to find more viable Big Data analytic and storage solutions. Enterprises engaging in the video surveillance arena must understand customer requirements and market trends to give full play to their innate advantages. These enterprises must team with IT companies to step into the Big Data era early on. Further, law enforcement and public security agencies must allow those with the know-how to provide the solutions and services to bring in the best practices from each industry and supply the best services to citizens.
The needs of society are the most compelling forces in technological development. Safe City and Smart City rollouts the world over are deploying the largest, multi-faceted video surveillance network systems even seen. This hotbed of activity provides an excellent platform for video surveillance and IT companies to learn from related IT technologies, designs, deployment, and operations practices built with this technology.
Practice makes perfect, as the saying goes, and practice helps improve theory. Applying IT to video surveillance will help drive IT technology in general. Through learning from the experiences of others, and with the concerted efforts of all stakeholders, the levels of productivity and positive output from IT going into video surveillance will go even higher.