ISS SecurOS Person, Vehicle, And Animal Tracking Analytics Test

By Rob Kilpatrick, Published Jul 14, 2021, 05:45pm EDT

ISS offers animal detection in addition to the more commonplace person and vehicle detection in their SecurOS Tracking Kit analytics. But how well does it work in real-world scenes?

IPVM Image

We tested the ISS person, vehicle, and animal detection examining:

  • Does ISS accurately detect people, vehicles, and animals in indoor and outdoor scenes?
  • Are all of these objects accurately classified?
  • Are people detected running, crawling, obstructed?
  • Does it ignore common outdoor false alarm sources such as foliage, shadows, lights, and rain?
  • Does it ignore common indoor false alarms sources such as lights and shadows?
  • Are crowds, loitering, line crossing, and direction accurately detected?
  • How well does forensic search (color, size, object type) work and what limitations does it have?
  • What PPF is required for accurate person detection?

*******

** ~* ***** ** testing, ***’ ******* ******** Kit ********** ******** *** classified *** ******, ********, and ******* (*****, ****, dogs), **** ***** ******** accurately ******** **** ******** or ********* **********, *** specific ****** ** ******* subjects.

*** *** ********** ***** alert ****** **** ***** droplets ** ***** ********** both ****** *** ******* false ********** *** ****** detection ********* ***** ****** on ******** ** ** fewer ****** **** ********** in ***** *****.

*******, ***** *** ****** forensic ****** (****** ****, color, *** ****), **** a ****** ****** ** searchable ** * ****, unlike ****** **** ****** across ******** *******, *** color ****** *** ********** in *** *******, ********* objects **** ****** ***** than ***** ********.

Vs. *********** *********

***** ** *** ***** alert **********, ********* ********, and *** ******** *** detection, ***' ******* ******** Kit *********** ** ***** average ******** ** *********** analytics ********* ***********'* ******* * **. *******, ***** ****** on **** ******** *** not ***** ** *******-********** *********** *** *****, **** as********'* *********' **** ******** ****** Analytics.

************, *** ****** **** capabilities ***** ****** ** not, **** ** ******** alerts ** ******* ****** and ********* ** *******.

*******, ******* ******** *** events ****** ** ********** into *****-***** ***. ********* are ** ******** ****** built **** ***' *** VMS.

People ******** ********, *******, **********

** *** *******, ******* accurately ******** *** ********** walking, ********, *** ********* obscured ***** ********:

IPVM Image

************, ******* ******** * distinct "*******" ***** ** humans, ************** **** ***** moving ** ****** ******.

IPVM Image

False ****** *** ******* ****** ** **** ********

*** **** ********** ***** alert ****** ** *** ISS ******* *** ******** collected ** *******' ***** triggering ***** ****** *** vehicle ******. ** **** cases, ***** ****** **** triggered ** *** ********** of *** ******'* *** IR ** *** ********.

IPVM Image

Update: ** ***** ****** ** **** **** "********** ****** ****** ** ******" ** *******

** *** ******** *******, due ** ** ************ error *** ********** ****** filter ** ****** *** not ******** ******* ***** resulted ** **** ***** alerts ** ***** ****/**** drops ** ****. **** proper ************ *** *** additional ****** ****** ******* there **** ** ***** alerts ****** ***** ****.

No ***** ****** ** ******* ******* ** *******

***** ****** ********* ******* false ***** ******* **** as ******* ******* *** not ******* *** ***** alerts, ****** ** ****** or ********:

IPVM Image

*** **** *** **** of ******* ****** ** blowing *******:

IPVM Image

No ***************** ** *******

*******, ***** *** ***** animals ** *** *****, e.g., *** ***** *** small **** *****, **** not ********** ** ****** or ******** ** *** tests.

IPVM Image

Accurate ****** *********

*** ******* ******** ******** an ****** ** ***** on ******* ** *** scene ***** ********** ******** dogs, ****, *** ***** in *** *****.

IPVM Image

*******, **** **** ****** detection ** * ******** analytic **** ******/******** *** SecurOS ****** **** * single ******* ** ** run ** **** ******. Because ** ****, ***** must ****** ** ****** humans *** *****************, *** ****** ****** all ***** **************.

No ***** ****** ** ****** ****** *******

****** ****** ******* ** false ****** **** ******* by ******* ********, **** as ***** ***** ******* (lights ***** ****** **/***).

IPVM Image

************, ******* **** ****** in *** ***** ******** by ****** ******* **** not ******* **.

IPVM Image

Other *********: ******** **** ********, *********, ***** *********

*** ******* ******** *** includes ********** *********, **** as **** ********, *********, and ***** ********* *********, which *** ***** ** their ****** *** ******* detection.

** *** *******, ******** and ****** ******** ***** were ********** ********, **** no ****** ** ********* detections.

IPVM Image

********* ********** ******* **** a ****** ******** ** the ***** ****** * preset ****:

IPVM Image

*******, ***** ********* ****** were ********* ********** **** a ****** ** ******* traveled *** ***** *** in *** *****, **** as *** *** ********* to *** **** *****, instead ** ** *** right.

IPVM Image

False ***** ********* ******

******* ******** ***** *********, as ****, *** ** our *****, ** ***** frequent ***** ****** ** vehicles ******** *** *****:

IPVM Image

************, *** ******** ********** alerted **** ***** ****** than ********** **** ** the *****. *** *******, despite *** ******** ***** set ** ***** **** on ****** ** * or ****, ** ****** on *** ****** ** the ******* *****.

IPVM Image

Update: ***** ****** ******** **** "********** ****** ****** ** ******" ** *******

** *** ******** *******, the ********** ****** ****** of ****** *** *** properly ********* ***** ******** in ***** ****** ** vehicle *** ***** ***** alerts ** **** ****** than *** ** *** threshold. **** *** ********** neural ****** ********* *** working ********, ***** **** no ***** ***** ****** on ******** ** *** scene *** *** ****** amount ** ****** **** alerted **. *** ******* the ********* *** ***** alerts ** *** ** three ****** ** *** below ***** *** ********** alerted **** ***** ****** were ** *** *****.

IPVM Image

Forensic ****** **** ********* ** *** ******

*** ****** ******** ******, but ****** *********** **** as******** ********** ****************'* ******* ********** ******, ***** ***** *** user ** ****** ****** multiple *******, ******* *** only ****** * ****** camera.

IPVM Image

*** ******** ** **** issue:

****** ****** ******** ******* (for *** ******* ******** Kit ******) ** * known ******* *** **** be ********* ** ** upcoming *******.

Searching ** ***** ********* **** ***-******** *****

********* ** ***** *** inaccurate *** ***** **** subjects *** ******* *** specified *****. *** *******, while ** ***** *** subjects ** ******, *****, it **** ******** ******** wearing ***** ****** *** blue *****.

IPVM Image

*********, ******* ***** ***** all ******** ** *** selected *****, *** **** found *** ******* ***** partially ********* **** *****. For *******, ********* *** black ******** ****** * van *** ***** **** black ****, ****** **** were ******* ***** ******.

IPVM Image

*** ********:

*** ****** ********* ******* 2 ****** *** ****** by ******. ** ** the ****** *** ****** one ** *** ******, it **** **** ** in *** ******.

Reliable ****** ********* ** ~** ***

****** **** ******** ******** at ~** ***, *********** with ***** ********* **** as ******** *** ********. This ******** ** *** following ********* ***** ***** a ***** ******.

IPVM Image

*****

***** *** ****** *** only ******** *** **** to ****** ***** ********* analytic ** *** *** draw * ****** ** interest, ***** *****.

Versions ****

*** ********* ******** **** used ****** *******.

  • *******: **.*.***

Comments (4)

** ***** **** * very ****** ******** ** AI **********, * ***** like ** **** *** know *** ** ******* with *** ******* ** vehicle ******* ******

Agree
Disagree
Informative
Unhelpful
Funny

*****, ** **** * test ** *** *** LPR/Car **** *** ***** detection ****:

*** ******* **** *** And *** ****/***** *********** Tested

Agree
Disagree
Informative
Unhelpful
Funny

*****,

** *** ***** **** to *** *** ******** in ****** *** ******** please ******* ** *** I **** ******* * demo *** ***.

**** *******,

******* ** *****

Agree
Disagree
Informative
Unhelpful
Funny

******

** *** ******** ******* the "********** ****** ****** of ******" *** *** properly ******* *** ** an ************ *****.

IPVM Image

**** ******** ** ***** alerts ****** ***** **** and ***** ***** ********* alerts ** ******** *** a ***** ****** ** people **** *** *********. After ******* **** *** additional ****** ****** ** people ******* ******** ***** were ** ***** ****** during ***** ****, ***** crowd ********** ** ********, and ***** ****** ********** detected ** ****** ** people ****** **** *** set *********. * ***** an ****** ** **** section ********** *** *** results.

************, * ***** *** feedback ** *** ******** color ****** ***** ***** was *** *********:

*** ****** ********* ******* 2 ****** *** ****** by ******. ** ** the ****** *** ****** one ** *** ******, it **** **** ** in *** ******.

*** ******** *** **** added ** *** "******** search **** ********* ** a ****** ******" ***** with *** *********:

****** ****** ******** ******* (for *** ******* ******** Kit ******) ** * known ******* *** **** be ********* ** ** upcoming *******.

Agree: 1
Disagree
Informative: 1
Unhelpful
Funny
Read this IPVM report for free.

This article is part of IPVM's 7,214 reports and 961 tests and is only available to subscribers. To get a one-time preview of our work, enter your work email to access the full article.

Already a subscriber? Login here | Join now
Loading Related Reports