Detektor International No 2

Special Feature: Intelligent Security Systems

Deep learning takes machine learning

to another level based on neural net-

work principles that impersonate the com-

plexity of the human brain."

1 2 · d e t e k t o r i n t e r n a t i o n a l

Security News Every Day www.securityworldmarket.com

More and more data are gathered by cameras and other security products and
video analytics and algorithms are becoming smarter and more efficient. This leads
to a point where the data can be used to streamline end customers' operations.

Intelligent security systems provide business opportunities

According to the market research

company Omdia, the market

for video surveillance is recover-

ing quickly after the suppressed

demand in 2020 due to the pan-

demic and the revenues will grow

to 31.9 billion dollars by 2025

with a total CAGR of 7.1 per cent

between 2020-2025.

Omdia expects average selling

prices of video surveillance equip-

ment to increase in the short term

because of US-China geopo-

litical tension and supply chain

constraints from the Covid-19

pandemic increases pressure of

vendor margins. However, in the

long term, higher average prices of

network cameras are also expected,

because of a greater adoption of

the network camera products us-

ing AI deep learning acceleration

and inference at the endpoint.

Oliver Philippou, Research

Manager, Physical Security at

Omdia says: "The semiconduc-

tor shortfall driven by geopolitics

and Covid impact has affected all

industries, and video surveillance

is no exception. This disparity

between supply and demand is

expected to last until mid-2022."

Increased efficiency
Another market research company,

Marketsandmarkets, emphasises

that the growth for edge-based

video analytics will be greater than

for server-based video analytics.

New powerful in-built chipsets

in cameras offer higher computa-

tional capabilities at the edge and

more sophisticated video analytics

such as queue management and

heat maps offer new opportunities

for businesses to increase sales and

improve traffic intelligence, ac-

cording to Marketsandmarkets.

Advancements in deep learning

and its integration with the edge

system are also expected to drive

the adoption in the coming years.

Marketsandmarkets writes: "Deep

learning takes machine learning

to another level based on neural

network principles that imperson-

ate the complexity of the human

brain. Earlier, the functionality

was mainly available at server-side

processes, which would require

videos to be decompressed and

processed. Edge-based devices

need external inputs to learn from

before proving as a useful tool

to recognise known objects and

behaviours".

IoT and smart city projects
Marketsandmarkets believes the

rise of smart cities and IoT are

expected to provide future growth

opportunities for the video surveil-

lance market since "modernised

cities require innovative solutions

to overcome long-term security

challenges". The company suggests

that IP cameras with increased

hardware functionality and

utilisation of AI analytics in video

monitoring systems are helping to

deal with the increase in secu-

rity threats worldwide. However,

privacy and security concerns over

video data restrain the market

growth.

The APAC market is expected

to account for the largest share of

the video surveillance market dur-

ing the next five years. Growing

population density in major cities

and investments in smart city

projects are two reasons for this

and the increasing number of retail

chains, small businesses, residen-

tial apartments, and hospitality

businesses, as well as government

initiatives in various countries to

install video surveillance cameras

in public places, such as cities,

schools, colleges, religious places,

and airports, is driving the market

growth.

Marketsandmarkets believes the

increasing population in develop-

ing countries such as China and

India, as well as decreasing costs

of video surveillance equipment,

is creating significant growth for

smart video surveillance solutions.

The use of data
Anders Karlsson, Bosch Building

Technologies Product Man-

ager for security products in the

Nordics, thinks we will see a large

increase in intelligent algorithms

that will be available in the

market. And he believes that the

purpose of many of them will

be to manage organisation and

optimise the operations.

He says: "With traceability

and a proactive flow of informa-

tion, you will know if you need

to reduce staff on a certain day

or add staff to increase sales.

This generates cost savings and in-

creased revenue and the cameras

bring added value."

Until now, Anders Karlsson

believes that the algorithms have

not been good enough and that

a lot of competence and data is

required to succeed. For example,

in retail, the algorithms have had

problems with people using shop-

ping carts and also with assessing

age and gender. In order for video

analysis of business data gathered

from security cameras to really

make an impact, he believes that

the knowledge must increase and

industries must meet and find

new forms of cooperation.

Implementing continuous

measurements of flows and analy-

sis contribute to streamlining the

end customers' operations. But

many people are still relatively

unaware of what can be done

with the help of analysis of data

collected from security cameras

and other security products.

Christopher Åkesson, Sales

Manager Nordics for Milestone

Systems, states that there are