Accenture video-analytics-operational-marketing-and-security-insights-from-cctv

Video AnalyticsOperational, Marketing and Security Insights from CCTV

Organizations now have an opportunity to increase the returns they receive from their existing video surveillance infrastructures. By applying sophisticated computer vision algorithms to video feeds, and adding an analytics layer to existing video surveillance systems, CCTV’s full potential can begin to be realized.

1

Enhanced security capabilities are just part of the picture. For example, retailers and customer-facing branch network operators are using insights from video analytics to optimize their operations and better understand customer behaviors. Municipalities and local governments are increasing situational awareness for intelligent cities.

And airport, train stations and mass transit operators are facilitating people flows, detecting operational incidents, and using predictive modeling to optimize their operations.

The availability of advanced video analytics technologies means it is time to take video surveillance to the next level—in security, commercial and public service environments.

Fuelled by concerns over terrorism and rising crime rates, video surveillance installations have surged in recent years. A study by research group RNCOS estimates that the global market for CCTV has been expanding at a CAGR of 20.5 percent since 2012 and will reach around US$23.5 billion by the end of 20141. While London has more cameras deployed than any other major city, most other industrialized countries have also adopted this technology. Looking ahead, China is predicted to become the world leader in this field, accounting for 70 percent of all video camera shipments by 20142.

The rapid increase in CCTV installations means that existing teams of operators are unable to monitor the vast quantity of

video data that is being generated. In what is a textbook example of data velocity4 overwhelming humans, estimates suggest that up to 98 percent of CCTV footage remains unseen5. As a result, most video surveillance installations are only used for forensic and evidential purposes after the fact.

But with the introduction of video analytics, most, if not all of that video data can be more effectively scrutinized, leading to major improvements in 24/7 video surveillance capabilities. What is more, the same technologies can also be used outside the security domain to unlock exciting new operations and marketing opportunities for organizations across all industries.

“When thousands of cameras run 24/7 in cities like Boston, New York and Washington, D.C., video surveillance quickly becomes a big data challenge. Analytics and automation technologies are the only answer3.”

Realizing the potential with video analyticsUsing video analytics, organizations no longer have to rely solely on human operators to process and analyze large volumes of real-time or historical video data. Instead, they can tap into automatic surveillance and analysis of video streams (see Figure 1).

Figure 1: Potential of video analytics

From Video SurveillanceOperator monitoring 20+ screensShowing feeds from 100+ camerasTrying to stay alert!

Operations, marketing and security insights

To Operations Center

2

Video surveillance—a booming market

Video Management System

CCTV Cameras

Video Analytics

Video Management System

CCTV Cameras

There are a range of video analytics technologies now available (see Figure 2), from basic motion detection to more advanced capabilities such as counting, tracking, anomaly detection and complex behavioral analysis. By using advanced operational analytics—a rule-based type of data analytics that analyzes individual video analytics events to detect pre-defined scenarios—actionable insights can be generated from the information received.

The benefits that this can have in the field of security were thrown into stark relief

following the April 2013 Boston bombing, when security agencies used video analytics technologies to scrutinize the thousands of hours of CCTV video footage that were available6. In another example, the UK and Netherlands border agencies have deployed an Accenture-provided self-service passport control solution (incorporating automated face capture and recognition) to process millions of passengers each year. Using self-service gates and face recognition technologies to read ePassports, validate documents, and carry out face capture and matching, passengers

Figure 2: The art of the possible: A sample of the rich catalogue of video Analytics functionality

3

can cross the border in a self-service fashion within, on average, eight seconds.

These are among the more well-publicized examples of the benefits of video analytics, and their successes are leading organizations across all industries to invest in these technologies. According to one recent report7, global “Intelligent Video Surveillance (IVS) & Video Analytics (VA)” industry revenues totaled US$13.5 billion in 2012, and are estimated to reach US$39 billion in 2020.

People Recognition Crowd Counting

Post Event Analysis

Traffic Monitoring

Video Indexing

Object Recognition

Safety Alerts Cross-Camera Tracking Footfall Tracking

License Plate Recognition Incident Detection Left Object Detection

Augmented Reality Wide Spectrum Imagery Suspicious Behavior

Breakthrough advances in technologyThis investment is accelerating in step with breakthrough advances in this area. New technologies such as IP storage are transforming how CCTV data can be collected, stored and used. Cloud-based video surveillance—“video surveillance as a service” (VSaaS)—is set to take the flexibility of CCTV installations to a new level. And big data technologies can now make unstructured CCTV data searchable and available for analytic processes.

As CCTV goes digital, with major improvements in optics, resolution and frame rates, the quality of data is rising all the time. Crucially, this data can now be seamlessly integrated into the IT infrastructure, where developments in IP networks, bandwidth, processing and storage mean that it can be stored and combined) with other enterprise data to enable insights that would previously have been impossible.

With increasingly sophisticated algorithms now available, the accuracy of these insights is improving. New face recognition algorithms in 2006 were 10 times more accurate than those in 2002, and a hundred times more accurate than those in 19958. People-counting algorithms have also been maturing at pace. Analyzing the shape of a living body (typically heads and shoulders) through video pictures, these can now count the number of people coming in or out of set areas with accuracy above 95 percent9.

Crucially too, analytics systems are not just getting more “intelligent”. They are also becoming more affordable, as the cost of image processors and communications systems continues to fall.

Extending the value of video analytics As video analytics technologies continue to evolve and become more affordable, organizations are beginning to investigate how they can be used to deliver heightened visibility into operational performance, as well as what is happening in the external environment. This has multiple benefits. For example, by being able to count the number of people within a facility and measure dwell time and crowd density—where people are congregating and for how long—an organization will know where to optimize customer flows and how to target marketing resources to maximum effect. And because operations centers are alerted when crowd numbers approach pre-defined health and safety thresholds, they know when to start restricting entrance.

Key information on physical space usage, footfall, wait and service times and typical usage patterns is all within reach in a real-time, continuous, 24x7 fashion—rather than through occasional manual observation surveys.

A recent Accenture pilot illustrates the benefits that this delivers in the commercial sphere. An Accenture core video analytics solution, which includes counting and queue monitoring capabilities, has recently been tested in South Africa. It is providing operational, marketing and security insights from CCTV video feeds, monitoring people and vehicle flows, queues and waiting times in customer service environments. In another example, Accenture’s face recognition solution has been successfully used in a large European shopping center to scan up to 15,000 customers per day, and generate notifications for known individuals with less than 0.5 percent false alerts.

Under pressure to do more with less, municipalities and local governments are embracing video analytics solutions that can optimize utilization of limited resources while maintaining day-to-day city operations and improving emergency handling. The Singapore Safe City Pilot is a prime example. Along with its industry partners, the Singapore government is piloting the integration of advanced analytics into existing sensors and systems across the city. The goal is to maximize situational awareness, streamline operations and enhance response capabilities across several different government agencies10.

Since April 2013, Accenture has deployed its Video Analytics Service Platform in Singapore to connect to existing and new sensor infrastructures (including dozens of CCTV cameras), apply computer vision and predictive analytics to surveillance video feeds to detect various events, and generate business alerts for six different government agencies.

4

Why Accenture?Building on its strong credentials in Biometrics and Facial Recognition Technology (FRT), Accenture has developed a layered, modular, vendor agnostic solution for video analytics that enables organizations to unlock greater value from their existing CCTV investments.

Illustrated in Figure 3, the Accenture Video Analytics offering applies computer vision algorithms to video feeds, adding an analytics layer to existing video surveillance systems to detect events of interest. By applying streaming analytics, the Accenture solution turns these events into business alerts that are displayed on an advanced touch-wall user interface.

Developed by Accenture’s Emerging Technology Innovation teams in Sophia Antipolis, France and Bangalore, this software solution is now delivering results for security services, public service agencies and commercial organizations worldwide.

5

Figure 3: Accenture Video Analytics: A layered, modular solution

Visualization

Core Analytics

Data Sources

Video AnalyticsVideo AnalyticsSocial Media Listening

Sensor Monitoring

Operational Systems Monitoring

Face Matching

CCTV Cameras People 3rd Party Sensors3rd Party Operational Systems

Traffic Monitoring

Anomaly Detection

License Plate Matching

Tracking

People Counting

Crowd Detection

Flow Management

Vehicle Classification

Operations External Systems

Operational and Predictive Analytics

The digitalization of video surveillance data, combined with wider availability of advanced video analytics technologies, means that more organizations are turning to these solutions to create intelligent business operations.

From automated passport control gates to next-generation applications in safe cities, video analytics is already in use and providing tangible security, business and customer service benefits.

As well as applying video analytics to video stream data to obtain deeper, richer operational insights, organizations are taking their situational awareness to the next level and strategically using these technologies to “engage with customers and citizens in new ways.

6

Video analytics: Toward intelligent business operations

About AccentureAccenture is a global management consulting, technology services and outsourcing company, with approximately 261,000 people serving clients in more than 120 countries. Combining unparalleled experience, comprehensive capabilities across all industries and business functions, and extensive research on the world’s most successful companies, Accenture collaborates with clients to help them become high-performance businesses and governments. The company generated net revenues of US$27.9 billion for the fiscal year ended Aug. 31, 2012. Its home page is www.accenture.com.

About Accenture Emerging Technology InnovationAccenture’s Emerging Technology Innovation group is responsible for incubating the next generation of Accenture’s technology services. By partnering with Accenture Technology Labs, Emerging Technology Delivery and the Technology Industry Ecosystem, it provides expert services to identify, pursue and deliver advanced and emerging technologies with pioneering clients.

Contacts To find out how Accenture can help your organization realize the full potential of your CCTV investments, contact:

Cyrille Bataller Managing Director - Emerging Technology Innovation [email protected]

www.accenture.com/videoanalytics

References1. www.rncos.com/Market-Analysis-Reports/Global-CCTV-Market-Forecast-to-2014-IM374.htm

2. ‘Eyes Everywhere’, pub.2011, Routledge

3. http://www.informationweek.com/government/information-management/whats-next-in-video-surveillance/240155161

4. www.accenture.com/technologyvision

5. http://icetana.com/product/overview/

6. http://gcn.com/Articles/2013/04/18/How-video-analytics-reconstruct-Boston-Marathon-bombings.aspx?Page=3

7. http://www.dailymarkets.com/stock/2013/03/25/intelligent-video-surveillance-vca-va-market-worth-39-billion-in-2020-say-new-research-report-at-reportsnreports-com/

8. National Institute of Standards & Technology (www.nist.gov/index.html)

9. http://www.contingencytoday.com/companyinformation/Ipsotek-Ltd/1435

10. www.channelnewsasia.com/news/singapore/new-challenges-to/632142.html

Copyright © 2013 Accenture All rights reserved.

Accenture, its logo, and High Performance Delivered are trademarks of Accenture.