Video analytics to detect fighting and collapsing

Posted on 2013-12-27
Last Modified: 2014-02-10
Is there any robust video analytic solution on the market out there to detect people collapsing and fighting?

If yes, would like to know how they can detect.

Many thanks
Question by:trihoang
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
  • 2
  • 2
  • 2
  • +2
LVL 38

Expert Comment

by:Rich Rumble
ID: 39774922
I know there are "downing" cameras, they monitor pools for people who don't seem to be moving any longer or are panicking. I'm not sure about fighting or other distress, nor how good it would be because it seems likely to false positive. Someone tripping might trigger it, someone going to tie their shoe quickly could. Fighting could be troublesome in certain places, but in others could probably be detected easily.
I know there is research on it but I'm not sure about actual software, but lots of good references in the documents below.¿ 2009 study of such software in "Transit" facilities.
LVL 47

Expert Comment

ID: 39774944
Yes, but you better have millions of dollars budgeted to productize it for civilian use.  If you don't then just forget it. What you ask for takes a heck of a lot of horsepower and the software isn't available commercially.
LVL 44

Expert Comment

ID: 39774963
>>Is there any robust video analytic solution on the market out there to detect people collapsing and fighting?

Simple and reliable (and probably cheapest as well with current level of technology) - a person monitoring the camera feed.
Technology Partners: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

LVL 47

Expert Comment

ID: 39775032
agreed with Andy -- put it this way, a supercomputer and software costing millions of dollars isn't any better at telling the difference between pictures of cats and dogs than a 3-year old child.
LVL 38

Expert Comment

by:Rich Rumble
ID: 39775043
Not according to the studies I've linked/read, it's getting closer and some software does exist for the fighting aspect, but like I said it's prone to missing and or misidentifying. It's just not widely pursued. The same thing with the drowning cameras we have at our local pools, they only have to really look at someone doing a "dead-mans float" to false positive. But you'd rather have that alert than not if it's being missed. As long as the FP's aren't too high, the technology and cost could be reasonable, however you have to think that stabbing or shooting would be missed by such tech unless it sees a melee first.

LVL 44

Expert Comment

ID: 39775066
Stationary people are one thing, fighting is very different.
LVL 70

Accepted Solution

Merete earned 300 total points
ID: 39775518
Hi trihoang
This is quite a complex feature for current CCTV/Video surveillance
and is still pretty new.
Privacy is at war with this technology in regards to data protection.
Look in to>
Video Content Analysis or VCA.
Video Analytics, also referred to as Video Content Analysis (VCA), is a generic term used to describe computerized processing and analysis of video streams.

Adding value to your video surveillance system
VCA Technology Video Analytics offers a highly accurate, easy to implement and even easier to operate method of obtaining valuable management information from live or recorded video.
PDF>Applications of Video-Content Analysis and Retrieval

ABOUT it from Wiki
Technological developments

Computer controlled analytics and identification
Today’s High-definition CCTV cameras have many computer controlled technologies that allow them to identify, track, and categorize objects in their field of view.

Video Content Analysis (VCA) is the capability of automatically analyzing video to detect and determine temporal events not based on a single image.
As such, it can be seen as the automated equivalent of the biological visual cortex.

A system using VCA can recognize changes in the environment and even identify and compare objects in the database using size, speed, and sometimes colour.
The camera’s actions can be programmed based on what it is “seeing”.
For example; an alarm can be issued if an object has moved in a certain area, or if a painting is missing from a wall, or if a smoke or fire is detected, or if running people are detected, or if fallen people are detected and if someone has spray painted the lens, as well as video loss, lens cover, defocuss and other so called camera tampering events.

VCA analytics can also be used to detect unusual patterns in a videos environment.
The system can be set to detect anomalies in a crowd of people, for instance a person moving in the opposite direction in airports where passengers are only supposed to walk in one direction out of a plane or in a subway where people are not supposed to exit through the entrances.

VCA also has the ability to track people on a map by calculating their position from the images.
It is then possible to link many cameras and track a person through an entire building or area. This can allow a person to be followed without having to analyze many hours of film. Currently the cameras have difficulty identifying individuals from video alone, but if connected to a key-card system, identities can be established and displayed as a tag over their heads on the video.

There is also a significant difference in where the VCA technology is placed, either the data is being processed within the cameras (on the edge) or by a centralized server. Both technologies have their pros and cons.

Hopefully this should point you in the right direct.
Contact the appropriate organizations for more infomation.
I am not sure if this is available to just anyone and eitherway it would be very expensive with the complexities of acquiring that amount of data from the footage..
Regards Merete
LVL 63

Assisted Solution

btan earned 200 total points
ID: 39775555
In the same light of Merete, you are looking at Behavior Recognition technology or Video Content Analysis (VCA). You can catch this brief summary from below

Also you may find this interesting (e.g. see "Behavioural recognition technologies (BRT)" at pg 185-188 and the appendix of systems list

Normally, you will be looking at 4 main layers for visual analytics system as a whole
-visualization layer and interfaces which typically host the C&C and for operations viewing and monitoring
-business rules layer which dictates the workflow and business analytics using the selected analytics platform
-analytics algorithm layer which is supposed to be the "selective" layer but unlikely you be changing it often since the systems is supposed to be interoperable where use case deem fits
-video source or video media source which is the capture and source of the raw feeds into the upstream a/m, this is also critical mgmt system equipped with interfaces to the surveillance devices (inclusive in this layer)

Featured Post

Free Tool: Postgres Monitoring System

A PHP and Perl based system to collect and display usage statistics from PostgreSQL databases.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Anti-virus software today is fairly sophisticated, but virus writers are often a step ahead of the software, and new viruses are constantly being released that current anti-virus software cannot recognize. The key to anti-virus software is detect…
Article by: btan
The intent is not to repeat what many has know about Ransomware but more to join its dots of what is it, who are the victims, why it exists, when and how we respond on infection. Lastly, sum up in a glance to share such information with more to help…
Finds all prime numbers in a range requested and places them in a public primes() array. I've demostrated a template size of 30 (2 * 3 * 5) but larger templates can be built such 210  (2 * 3 * 5 * 7) or 2310  (2 * 3 * 5 * 7 * 11). The larger templa…

730 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question