The Ultimate Showdown: Video Analytics Viciously Defeats Motion Detection

Posted by Indigo Hansen on Feb 26, 2019 12:00:00 PM
Indigo Hansen

"Laaaaaaadies and gennnntlemen, welcome to the event of a lifetime. The showdown between two age-old rivals. A battle for the title. A match like no other.

In corner number one, we have the new kid on the block. The underdog. Viiiiiiiideo Analyyyyyyyytics.

In corner number two, we have the legend. The king of security. Mooooooootion Detectionnnnnnn."

 

Turing Blog - Let's Get Ready to Rumble

 

Let's take a look at the match.

SPOILER ALERT: In this showdown, video analytics (VA) whoops motion detection (MD) with a TKO in the first round. Read on to learn why.

 

The match was over before it began

Video analytics has emerged as a major security trend over the past 15-20 years. It's grown with hardware improvements, the proliferation of AI software, and a variety of other technological advancements.

Here are the reasons video analytics takes the title:

  1. Actionable intelligence
  2. Gets smarter with time
  3. Scalability and flexibility

 

Turing Blog - Rocky Balboa Victory

 

Reason 1: Actionable intelligence

Pop quiz: which alert would be more helpful?

  1. ALERT: Activity detected on third floor.
  2. ALERT: An unknown person was detected with a backpack in the third floor conference room.

The second option provides clear, contextual alerts, whereas the first still requires further investigation.

Motion detection tells you something has happened, but it doesn't know the difference between a car and a dog or a person and a backpack.

Video analytics tells you what, where, and when events occur. We call this type of information actionable intelligence.

 

Video Analytics example - walking in park

 

Reason 2: Gets smarter with time

Video analytics are powered by artificial intelligence (AI). Compared to motion detection, VA software is significantly smarter and, similar to a fine wine, it gets better over time.

As the AI software analyzes more video, it gathers more training data which improves its accuracy dramatically, and thereby reducing false positives to near-zero.

To learn more about AI and how it works, check our blog post "What's the Difference Between AI, Machine Learning, Computer Vision, and Deep Learning?"
READ THE ARTICLE.

Reason 3: Scalability and flexibility

"Change is the only constant."

In a sense, businesses and organizations are like sharks: when they stop moving, they die. This constant evolution demands that tools, workflows, and cultures be flexible and adaptive.

 

Turing Blog - Swimming shark

 

Video analytics are run on software that can integrate with any camera and hardware. As cameras, DVRs, NVRs, and VMS brands change, agile software can continue to adapt with new brands and new technology.

It's a different story with motion detection, which is limited by its design and hardware.

In most cases, motion detection uses sensors to pick up movement of objects within a certain size and distance range. These sensors are limited by an object's distance and size. Sensors wear down with time and the hardware can quickly become obsolete as technology continues to advance.

 

Turing Blog - Motion sensor fail

Or sometimes, those sensors stop working altogether.

 

Customize, customize, customize

One of the major benefits of using VA software is its ability to be customized for specific use cases. You are able to create custom rules when certain events are detected. Here are some examples:

  • Notify me when someone enters the office lobby any time after hours.
  • Dispatch an officer if a person is loitering in the parking garage.
  • Send me an email alert when an emergency exit is blocked on the first floor.

This flexibility allows VA software to grow with a company and its needs. As a company expands from one office to fifteen global offices, the software can adapt to the different tools, environments, and needs for each location.

 

Turing Blog - Customize from iOS device

 

On the other hand, MD isn't quite as adaptive. Optimizing a motion detector's sensitivity for one type of object can inadvertently affect its ability to detect another object.

 

The wrap-up

Before AI was widespread technology and before software became king, motion detection was on the cutting edge of security technology. In some instances, it still holds its place in deterring crime (shoutout to motion-activated lighting).

 

Turing Blog - The times are a-changin'

 

However, the times are changing. Smart software has proven to be the future of security, and video analytics is just the beginning.

 

Topics: Automation, Smart Security, Emerging Technology, Artificial Intelligence, Technology, Technology Trends, Deep learning, Machine learning, Edge computing, Monitoring, Surveillance, Video Analytics