ISS SecurOS Auto LPR And Car Make/Model Recognition Tested

By Rob Kilpatrick, Published Jul 07, 2021, 01:29pm EDT

ISS' SecurOS Auto LPR includes vehicle make and model detection, a rare feature in competitive analytics, but how accurate is it in real world scenes?

IPVM Image

We tested ISS SecurOS Auto examining:

  • How accurate are license plate reads overall?
  • How well does it handle stacked letter or specialty plates?
  • Does it detect any other text on the vehicle?
  • Can it detect both front and back plates?
  • Is vehicle make/model detection accurate?
  • Can plates and make/model be identified in heavy rain?
  • Are colors accurately detected?
  • How SecurOS Auto compares to Genetec, Axis, and Flock Safety

*******

** **** * **** of *******, *** ******* Auto ******* ****** ***** and **** ******* ***** recognition **** ***** **** 10 ******** ****** ~*** license ****** ********** **** in **** ******* **********. Recognition *** ******** **** cars ******* ** ** ~65mph, *** ******* **** could ****, ****** ******* is ********* ** ** 155mph. ******* **** **** all ******* ********* *** specialty ****** ********* ** it, ** ***** ** IPVM ***** ** ***** LPR *********. *** ***** were ********** ******* (******** approaching *** *******) **** a ****** ***** ******.

*** ***** *** ****** were ********** ********** ****** ~500 ****, ******, *** SUVs, ********* ** **** weather **********. *******, ******** such ** **** ****, garbage ******, *** ******, and ***********, ***., **** misidentified ** *** ********** at ***. ****/***** *********** worked **** ** ***** horizontal ******, ******** * license ***** ** ******** to ******* ***********. *** color ********* *** ******* identifying **** ****** *********, such ** *****/****/******, **** in ******* ******** **********.

*** ******* **** *** up *** *********** ******** to *********** **** ****** require ******** *** ********** or *** **********, **** optional *******/******* ***** **** be ********* *** ******** in ******** ** *** VMS, *** ******* ******* which ******** ************* ** eliminate ***-******* ***** ***** detections **** ** ***** lettering.

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

***' ******* ********* ***********'* ****** ***** ***** ********* ** make/model *********. ******* ****** only ******* **** *********, though ** ***** ** was *** ** ******** in *** *******, *** frequently ****** ******* ****.

***** *** ******** *** similar, **** **** ********** reading ******** ******, ******* characters, *** ********* ****** at ****** ** ** ~65mph, **** *** ****** reading ******* ********* ** missing ****** ********.

** *** ******** ** updated **** ** ****** in *** **** *** months.

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

*** *** **** ******** than****' ******* ***** ********, ***** ******** **** numerous "****" ***** ** non-plate **** *** ***** objects. ************, **** ** intended **** *** *** speed *******, ** ** 45mph, ****** ***' ****** specs, *** ** **** not ******* ******* ****/***** detection.

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

*** ******* **** *** performed ********* ******* ****** ***, **** **** ********** reading ******* ********** *** specialty ******. *******, ****** ISS, ***** **** *** support ***** ***** *****. Lettering ** ******** ***** than ******* ****** **** not **** ** ****** ISS *** *****.

*** ********* ****** **** identifying *** ****/*****, ********** detecting **** *** ****, trucks, *** ***'* ***** Flock ****** ** ********** detect *** **** ** a *******. ****** ***** has *** ********* ** LPR ***** **** ****** a ***** ******* ****** while *** **** ** connected ** *** ******** on *** ******* **** server *** ***.

*******

*** ******* **** *** two ******* ***** *** low ***** *** **** speed. *** ***** **** is $*,*** *** ****** and ******** ****** ** to ***** (****/*), **** speed **** ** $*,*** per ****** *** ******** plates ** ** ****** (250km/h).

Complex *****/*************

***** *** **** ******** than **** *********, ********* installation *** ********* ** optional *** *******, ** well ** ******* ***** setup (*.*., *******/******* ********* heights) ** ********* ***** plate ************ ** ***-***** lettering. **** ** ***** tasks ******** *** ********** to ******** ****** *** tests.

**** ** *** ***** below ** *** ***** setup ** ********** *** plate ***** *** ****** non-plate *********.

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

**** *** ****** ** about *** **** ** testing, ** ***** **** plate *********** *** ********, identifying ******* *** ******* accurately ********* **** ********* characters **** ** *, X, *, *, ***.

IPVM Image

*** ******* **** **** properly ********** ***** ****** on ********, *** *** night.

IPVM Image

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

*** ******* **** ********** captured ****** ** *****.

IPVM Image

Plates ********** ** ***** ****

******* ****** **** ***** identified ****** **** ******* rain, ******* ***** ***** spray ********* ***** *******, shown *****.

IPVM Image

Fewer **** ** ******** ****** ~*** ***** *****

****** ~*** ***** ***** there **** ***** **** 10 ********, **** *** most ****** ***** ***** plates ******** ** ******* or ****** ****, **** very *** ******** ********.

IPVM Image

*** ******** *** ** adjust *** ******* ********* size ***** ***** **** but *** *** ********.

IPVM Image

Two **** ********

****** ******* ******* **** accurately ********** *** ***** of ******* ** * single ***** ******, ********* users ** *** ********* of ****** (***********/*******).

IPVM Image

Stacked ********* *** ********* ****** **** *********

******* ***** ********** **** read ********* ********** ** whether *** ******* ********** were ** *** **** or *****.

IPVM Image

Vehicle ****/***** ********

***'* ******* **** *** model ********* ********** ********** hundreds ** ******** ****** ~2 ***** ** *******. Note **** ****/***** *** detected **** **** * plate ** ****, *** simply ** ******* *********.

IPVM Image

Cars ********** **** *****, ****, *** ****

******* **** ******** ********** cars *******, *********** ** from *** **** ****** a ******* ***** ** needed ** *** ***** to ******** *** *** make/model.

IPVM Image

Vehicle **** *** ***** ********** ** ***** ****

****** ***** **** ******* make *** ***** **** accurately **********.

IPVM Image

Work ******, ********, **** ******, *** ********** ** *********** **********

****** ****** **** ** garbage ****** ** **** trucks **** *** ********** from *** **** ** incorrectly ********** ****** * license ***** *** ***** read.

IPVM Image

Motorcycles *** **********

*********** ******** *** ***** had ***** ******* ****** identified *** *** ********** not ********** ********* ** a "****** ** ******** the *******" *******.

IPVM Image

Non-Plate **** ******* ***** ****** *****

***** ******* ********, *** recognized ***-***** ******* **** as *** ******* ****** below:

IPVM Image

** ***** *********/*********, ***** here:

IPVM Image

***** ****** *** ** (adjusting *******/******* **** ******, shown ** *** ***** above) ***-***** **** *** ignored **** ** ******* or ***** ******* ** back ** **** ** a *******.

IPVM Image

Car ****/***** ******** ****** ** *****

** ***** ** **** accurate ***** ***** ** night *** *** *** not ******* ****** ** make ******** ****/***** ************ resulting ** ******** ** classify ********.

IPVM Image

Color ********** ******

******* **** *** ****** detecting *** ****** ******* color **** ** ****-*** scenes. **** ****** ******** on ***** ********, ***** were ***** ********** ** silver *** **** *****, or ****, ********** ** silver ** ***** ********* on ********.

IPVM Image

Versions ****

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

  • *******: **.*.***

Comments (13)

**'* ********** ** *** that ***** ***** *** excluded **** *** **********.

Agree
Disagree
Informative
Unhelpful
Funny

** **** *** ****** Rekor *** ** ** cannot ** * **** comparison ** **** ** have *** ******. ** are *** ******* ** testing ** *** ** have **** ******* ******** coverage ** ***** **** year, *.*.,***** ****** $** *******, Now $***+ ******* *********

Agree
Disagree
Informative: 2
Unhelpful
Funny

** ***** **** *** also **** ** * class ****** ******* *** securities ***** **** ****** they ****** *********. **** to **** ** **** a ********* ******* ** the ****** *** ***** or ***.

Agree
Disagree
Informative: 3
Unhelpful
Funny

******* ** *** *** allegations *** *****, * do *** ****. *** my ******* ********** ** that **** ******** ** not ********* ******* ** existential **** ** ********* even ** **** **** or ******.

Agree
Disagree
Informative: 1
Unhelpful
Funny

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

** *** **** *** press *******, ** ****** like *** *** **** is ***** ******* *** plaintiffs.

******* *****: ***** ******* Sued *** ********** *****; Investors ****** ******* ***** & ...

Agree
Disagree
Informative
Unhelpful
Funny

*** **** **** * solid ******** *** ********* reading, *** * **** I'm *** ** ** date ** *** ****** versions. ***** *** ***** a *** ******** ***** the ******** ** ******* specific ********** **** ********** to ***** ***** ** very **** ************ *** manufacturing ** ******** ********** functions.

Agree: 4
Disagree
Informative: 1
Unhelpful: 1
Funny: 1

******* *******. ***** ***. Yes, ***** **** ** see *** ***** ***** out ******* *** ******.

Agree: 2
Disagree
Informative: 1
Unhelpful
Funny

* **** ****** **** videos.

Agree
Disagree
Informative
Unhelpful
Funny

*** **** **** **** is ******* ** "******* Rate"

** ** ********* ** set ** * ******** device **** ****** ***** vehicle **** ****** *** compares *** ******* **** to *** *** ******* rate.

******** ** *** "********" until *** **** *** many ******** *** *** engine ******* ******* *** plate ** *******, *****, non-reflective, ***.

****, *** ** *** make/model/color ******* **** ** night? **** **** ********* are ***

Agree: 2
Disagree
Informative: 1
Unhelpful
Funny

****, ** *** ** capture **** ** *** not **** ********* ******** vehicles ** *** ***** but * ****** ****** footage *** ******* ** events *** * *** not ****** *** ****** day ** *****. *** road **** **** ****** is ******** ** ********** by ******* ****** *** dump ****** ******* ** is ** *** *** to * ******** *** they **** ** **** dirty, *******, ** ********** plates ***** *** ***** captured.

*** ******** *** ***** lettering ** *** ***** is ********* ********** *** still ******** (***********).

IPVM Image

**** ** ********* ***** the ***** ** **** partially ******* ** ** captured.

IPVM Image

IPVM Image

****, *** ** *** make/model/color ******* **** ** night? **** **** ********* are ***

** ***** ** ******* plates ** *****, **** were ****** *** ******* which ******** ** * failure ** ******* *** vehicle ****/*****, * ***** a *** ******* ** the ****** ** ******* this.

Agree
Disagree
Informative: 1
Unhelpful
Funny

*'** **** ***** *****, an ** ***** ****, at ** ********* *** almost *** *****.

****/*****/***** ***** **** **** sunrise ** ****, ** addition ** *****.** *****, it's ***** ******* ****.

*** ******* ***** ** the ******* ************ ***** route. * **** **** the ******* ** * red *** **** *** there ** *** ****** image **** ********* *** the *.*. ** ********* make/model/color **** ** ****, very ********, ** ********.

* ***** ***** ** A.I. **** *** ******* make/model/color **** * ****** shutter ************* *** ***** time ***** *******.

IPVM Image

Agree
Disagree
Informative: 1
Unhelpful
Funny

*** ***** *** **** only ****** *** *** or ** ***** ***?

Agree
Disagree
Informative
Unhelpful
Funny

***** *** * *** examples ** ********* *********** here, ****** *** ***** crop ***** ** **** difficult ** ***.

**** ******* ** *** right ** ** *** middle ** *** *****:

IPVM Image

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

IPVM Image

** *** **** **** larger ******** ** *** need ****, *** *** this *** *** *** night.

Agree: 2
Disagree
Informative: 2
Unhelpful
Funny
Read this IPVM report for free.

This article is part of IPVM's 7,216 reports and 962 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