Growing the Role of Analytics in Video Surveillance
Adding value to video surveillance
- By Charles McCready
- Feb 01, 2015
The situations that call for intelligent video, also referred to as video
content analysis and video analytics, are growing by the day. As
this technology continues to advance, on-board camera innovations
are unveiling a range of new features and functionality to
deliver additional value, as well as proving to be increasingly relevant
across a variety of vertical markets and customer applications beyond
traditional video surveillance.
Once reserved primarily for stable, high-security specifications only, video analytics
are growing into mainstream use for not only gathering security-related,
loss prevention information, but also to provide valuable data that can be passed
along to end-user customers in detailed reports. These reports deliver business
intelligence information that improve operational efficiency, deliver insights into
customer demographics that inform marketing strategies and ultimately improve
business profitability.
At the core of this innovation is greater image processing power in cameras,
leading to deeper, more targeted and easily deployable analytics. Advanced video
analytics record summaries of what the analytics scrutinize, such as size, object,
speed, trajectory, aspect ratio and/or license plate numbers. These advanced types
of video analytics applications can generate all kinds of information, known as
metadata, for future use. Camera-based, or edge analytics, are particularly advantageous
for remote locations with limited bandwidth, providing an effective solution
without requiring additional servers.
For the technology solution provider, increased processing power and memory
in the camera gives them the ability to apply analytics across a larger, more
diverse customer base. Now, providers can offer security cameras not only as a
way to provide traditional video surveillance and security, but also as a method
to extract additional data insights related to merchandising, marketing and even
inventory control.
The more the surveillance solution can yield measurable results and actionable
data, the more likely the technology provider will be valued as an integral part of
the company they are serving. The more data the end user can glean from their
cameras, the better the justification for additional spending and greater the likelihood
of using analytics for other applications. In addition, surveillance that moves
into the realm of business intelligence can be more easily justified as a regular business
operating expense rather than a capital outlay, more palatable to the end user
and their sometimes restricted budgets.
Evolution of Cameras
The evolution of security cameras has moved these devices away from the genre of
static, reactive security tools. They are now being used not only for surveillance and
to assess security at the protected premises in real-time, but have also gained the
capacity for analytic applications. The processing power is greater in today’s devices, and the addition of 4K innovation is having a positive
effect on analytics deployments. Ultra high-definition
4K cameras boast the ability to capture the highest
quality images ever, assisting in deeper analysis of surveillance
areas for more effective deployments and reduced
likelihood of false positives.
The growing availability of increasingly high resolution
cameras, such as a 12MP 360 camera that produces
a 9MP 3K x 3K fisheye image at 15fps and a
4MP fisheye image at 30fps, deliver more pixel-perfect
image detail across the broad field of view in a panoramic
format and allow the kind of detailed images
analytics thrive on for reliability and accuracy. In addition,
advanced image sensors in the latest cameras
are now capable of managing the light required to activate
the smaller, more numerous pixels in 4K images,
which also assists in cultivating analytics.
Applications across the Board
With 4K camera technology, advanced analytics can
help end users in multiple vertical markets obtain
ultra-high quality video to protect people, property
and assets, and provide additional tools for business
process improvement. Since some 4K cameras now
entering the market include analytics and facial recognition
capabilities, they are well suited for stadiums,
casino floors and smart city applications because of
their ability to zoom in and provide exceptional detail
on a particular individual, activity or potential threat
in large crowds.
In addition, 4K cameras with video motion detection/
analytics capabilities such as loitering and wrong
direction alerts are ideal for monitoring large, open
spaces such as airports, warehouses and public transit
terminals. In retail environments, heat mapping,
people counting and dwell times provide insights into
both personnel and customer paths, further uncovering
business optimization and marketing opportunities.
Advanced auto tracking, an intelligent video analytics
features combined with conventional video motion
detection is another innovation made possible by
higher processing power on board cameras. When an
advanced network camera finds a moving object in
its view, the camera automatically detects the moving
object displayed and triggers an alarm based on preconfigured
alert rules for loitering, direction, scene
change or other atypical activity.
Extracting Value from Video
In retail applications, the latest video surveillance
cameras are helping yield rich rewards beyond physical
security. Cameras with advanced analytics capabilities
provide more than video surveillance, protection
against shoplifters, and video evidence of
transactions, but other data collection information
that retailers are beginning to leverage to obtain a variety
of useful business metrics.
For example, these cameras can provide heat mapping
functionality that identifies which displays customers
linger at longest (dwell times), traffic patterns
of how customers move through stores and which areas
cause ‘pinch points’ where traffic becomes heavy.
When checkout lines or queues become long, store
management can be automatically alerted to add staff
at cash registers.
With new, field-proven image processing technology
and network cameras equipped with high-performance
processors, intelligent video analytics have
made substantial progress in extracting more information
and value from the field of view, such as approximate
age/gender, moving direction and loitering time
from the detected objects while analyzing behavior.
Analytics can help businesses improve their operations
with easy-to-interpret reporting. Additional
analytics include bi-directional “people counting”
function to assist in quantifying activity in a scene.
These new analytics are ideal for generating insight
into floor behavior in retail environments or to track
people movement in public spaces.
Other types of popular video analytics for indoor
application based on motion detection and enhanced
with advanced auto tracking include:
Loitering detection sends alerts of people who are
idling in the camera’s view. When a network camera detects
human-sized moving objects in a pre-configured
area, the camera starts tracking each of them. When
they loiter for a specific period of time, the camera
sends an alarm and highlights them with frames, helping
operators easily identify who is being tracked.
Direction detection detects persons, cars or moving
objects that go the wrong way down an aisle or street.
When the object is moving in an unauthorized or unusual
direction, the camera sends an alarm to the operators
and highlights the object the camera is tracking.
Face detection searches images for face-shaped
parts in real time and highlights them with frames,
providing the ability to watch those entering and leaving
a facility or protected premises.
With greater processing capability comes better
accuracy of deployed analytics and fewer false
alarms. Now, heat mapping is useful not only in retail
applications, but in premises where large groups
of people are congregating where perhaps they
shouldn’t be. Highly secure areas such as high-tech
biomedical facilities can use directional analytics to
manage the flow of traffic, and alert discrepancies of
traveled patterns.
The move to on-board analytics will continue to
open the market to many additional application possibilities,
including installations for small-to-medium
businesses (SMB). For technology solution providers
and SMB customers, cameras with on-board analytics
can be strategically placed in the protected premises
as a force multiplier and information extractor.
Camera surveillance is changing, with high-definition
images and advanced processing allowing for
embedded analytics that gather crisp images and add
a level of data that can be efficiently
communicated to the end user for effective
business reporting metrics.
This article originally appeared in the February 2015 issue of Security Today.