Vehicle Analytics (Color, Type, Make/Model) Rankings 2024 - Avigilon Alta, Axis, Genetec, Motorola, Rekor, Verkada

Avatar
Rob Kilpatrick
Published Feb 01, 2024 16:22 PM

Vehicle attribute classification is becoming more popular among many analytics systems, but how do they perform in real-world testing?

IPVM tested six different vehicle attribute classification offerings, including Alta LPR, Axis All-In-One LPR, Genetec Autovu LPR, Motorola L6Q, Rekor Edge Pro, and Verkada LPR. Flock has refused to sell us or allow us to obtain their products for the past two years.

**** ****** ******* *** ********* *** weakest *** *** ******** ******** **** the ********* **********.

  • ******* **** *** ***********
  • ******* **** ***** (**) ***********
  • ******* ***** *** ***********
  • ******* ***** *** ***** (*******, ****) Performance
  • ******* ***** ***** (**) ***********
  • ******* **** *** ***********
  • ******* **** ***** (**) ***********
  • ******* ***** *** ***********
  • ******* ***** ***** (**) ***********
  • **** ** *****
  • *************
  • ****** *********

***** *******, ** **** *** **** research,*** *** ** **** *************, ** ** *********** ***** ** ********** ** ******* *** *********** ********** of ************ *** ********, *** *********.

*** ******** ******* ******* **** ******** and ******** ******* ******** ***** ** test *******, **** * ******** ******* below:

IPVM Image

*** ****** ******** - ******** ****, ****, Genetec, ******, ********, *****, *******, ***** ******** ******* ********* *** compared *** *** ********* ** * real ***** *******.

Pricing **********

******* ****** ************* ******* ********* ********* include ***** ******** ************ ** ********* form *******, **** ***-** ***** ****** kits ** ******** **** *** ****** purchase **** ** ********** ********* ****.

IPVM Image

Our ******* ***********

** *** *******, ** ******** * month's ***** ** **** *** ******-******* all ******* ***** ** *** ** the ******** **** ****** *** ******* point. ****, ** ******** *** ****** of ******* ********* ************ ******* ********* recognitions. *** *** ***** ** ********* recognitions *** ******* * *********** *** filling **** */* ** *********** ** the ***** ****, ****/*****, ** *****.

** ********** *** ********** ** ******* recognitions ** *** ***** ****** ** recognitions, ** ********** *** ******** ** the ******* ********* *********, *.*., ** 100 ******** ****** *** **** *****, 100 **** ********, *** ** **** accurately **********, ** *** ******** ** 96% ********.

********* *** ******* ********, ** ******** what ************ ****** ****** *** ******* quantified *** ***** **** ***** ******** vs. *** ***** **** *** *** and **/*** ** ***** ******* ******.

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

***** ** * ***** ******** ** the ******* ********* ********, ******** **** criterion *** **** ** *** ****** we ********** ***** ******* ***** *******.

Executive *******

** ****'* *******, ***** ********* *** best ****** **** ********, ***** ******* consistently **************, **** ***** ************* ********** average *** ** ******* ******* ****** all ********.

******* **** ********* ****** ************* ****** than **** / *****. ** * follow-up ******, ** **** ****** *** most ****** ******** ********* *** *********** make / *****.

Summary ** ********

*** ***** ***** ***** *** ******* of *** ******* ********* ******* ********.

IPVM Image

****: ******** *** ****** ****, ****, and ***** ** *****, *** ** requires * ***** **** ** *** same ******* ****** *** ***; ******** will **** **** *** ********** **** the *** ***** *** ***** ** to *** *****.

**** *** ********** ********, ** ****** them **** *** ** **** ** performance, **** *** *** ** *** weakest *** **** ***** ** *** strongest.

IPVM Image

**** ********** ***** ***** ** ***** criteria *** **** ********* ** **** of *** ** **********. *** *******, if *** **** ** ****** ***** the ****** ** ******* ** **** at *****, **** *** **** ** choose *** ****** **** *** ******* night **** ******* **** ********.

Manufacturer ******* *********

***** *** ********* *** **** ** the ****** ******* ********* ************** *******, detailing *** ********* *** ********** ***** in *** *******.

Avigilon ****IPVM Image

******** **** ******** ******* **** *** color *********, ***** *** ******* *********** in *** *******. ******** **** ********** vehicles ** **** ***** ***** * general ******** ******* ** ********** ** classify **** ******* **** ************ ***** made *** ****** **** ******** ** the ******* ** * **** ********* search **** ****** ** **** ********** vehicles.

******** ****'* ******* ***** ** *** inability ** ******* **** **** *****, white, ** *** ***** ***** *****, often ****** ***** ***** *************** **** grey ** ******** ***.

***** *** ***** ** *** ******* analytics **** ** **** **** * clicks *** ******* ****** *** *******, Avigilon **** **** *** ******** ********* what ***** ** ****** * ******* is, **** ***** * ****** ** made *** * ******** ***** **** a ***** ************** ** *******.

*******, ******** **** ******** **** ************ events *** ******** ******* ****** *** classifications **** **** ******** ****** ~* seconds ** *** ******* ******* *** camera.

Axis ****** *********IPVM Image

**** ****** ********* *** ** ******* performer ** *** ******* ***** *** vehicle **** ********. *** ******* ***** was *** ~**% **** ** *********** when *********** ******* **** ** * dark (**) *******. **** ****** ********* is ******* ****** **** **** ********** vehicles *** ******, ******* *** *** pickups ***** *** "***" ************** *** will *** ** ******** * ****** shade ** *** ***** ** **** "Dark", ***** ***** ** * ****** perceived ******** *** ********* ****** **** a *** ** ******* **** **** searching *** ******** ******** ** ******.

**** ********* **** ******** ***** *** scope ** ***************, ****** *** ******* issue *** ********* ************** ******* ***** the "***" **************, ***** ** *** intentional.

*******, ***** **** ***** ****** * only ******** ***** ***** ****** * search. ****** ********* **** ********* **** a ***** ****** ** ***** *** Axis ********* *** ** ********** ** receive ****** **** **** ****** *********.

Genetec ****** ***********IPVM Image

******* ****** *********** *** *** *** issue **** ******** *** *********** *** that *** *** ********* ******** ******* metadata **** * ******* ******, ***** omitting *** ** *** ******.

***** ** ****** ******* *** ******* in ******* **** ********, ** *********** many ***** ** ******** ***** *** classification ** "********* *******", **** ***** include *****, ***, *** ***, **** a ******** ************** *** ****** ******. Since *** ****** ** ****** ********* for ***, ** *********** ***** *********** first, ***** ********* ***** ******** ** ~40% *** ~* ***** * *** during *** ******* *** ******* *****. Vehicle **** ******** ******* ** ~**% at ***** ***** **, ***** *** better **** *** ******** *** ***** provided ** ******* ****/***** ** **** at ***** ****** ** **** *** the **** ******* ***** ****** *** day.

*******, ******* ****** ******** ** **** than **** ******, *** ****** *** system ***** **** *************, ** **** provide ***** *************.

Motorola ***IPVM Image

*** ******** *** ********* *** ***** for ******* **** ******** *** ** does *** *** ** *** *** vehicle ***** ***** *** ******** *** instead ******** ** ******** ******** ******** such ** ****, ******, *** ****. This ******** **** **** ** **** less ******** *** ******** * **** detailed *** ** ****** ***** *** fact.

*** *** *** ******* *** *** performers **** *********** ******* ***** ****** the *** **** *** ******* ***** being **** ** ***** **** ******** for *********. ******** *** ******* ** Genetec *** ***** ** ** ****** optimized *** *** ***********, ******* ******** dropped ** ****/*******, *** *****.

*******, ******** ******** *** **** *********** with ****** *** ** *** **** options ** ****** **** *** ******. Push ************* **** **** ******** *** LPR ************* *** *** *** ******* any ******* ********, ** ** *********** only *** ******.

Rekor **** ***IPVM Image

***** *** *** *** ********* ******* many ******** **** *** ******* ***** being *** ~**% **** ** ***** recognition ** ***** (**). ***** *** the ******* ******* **** ******** *** was *** ** *** ******* **** did *** *** **** ***** ** vehicles ***** *** "***" ************** ***** made ****** ** ******* ******** **** likely ** **** *** **** ******* results.

***** ****** ******** ** **** **** 5 ******, ****** ************* *** **** provided *** *** ************ *** ******* metadata ** *********** **** *** ******.

*******IPVM Image

******* *** ** ******* ********* ** our ******* *** *** ******* ***** was *** ***** ***********, ***** ********** colors **** ***** ** **** ** red ** ****** ** ********* *** green ***** *** ******* ******* *** that *****.

******* *** * ****** **** ** hidden ** *** ***** ****** *** and *** *** ******* **** ****** the ****** ******* ******** ****. ******* also **** ****** ****** *** ****** and *** *******.

Lighting ******* ****** ** ***********

** *** *******, ******** *** *** biggest ************** **** ** **** ** accuracy *** **** ***** *** ******* make/model/type **************, ******** ******** ** ****** 20 ********** ****** ****** *** ******* when ***** **** *** ** ***** and ********** *** ****** ** ***** vehicle ****/****/***** ************.

Dusk/Dawn *** ***** ****** *** ******* *** ********

** *** *******, ******* *** ******** had ****** ***** ** *** ******* and **** ** *** *******, *** hour ****** *** *** ***** *** the *****. ***** *********** ******** ******* as **** ** ** ********** ****** on *** ******* *** ******** ******* with ** ****** ** ****** ******** to ****** **** *****.

IPVM Image

Color *********** *** ****** ** ***** (**) *******

*** ** ** ******* ** ** on ** *** ******* ****** ** night, *** ***** ******* ***** *** white, ********* ***** *********** ******* *** every ******* ** ********** ** *****, grey, ** *****.

IPVM Image

Sun ***** ***** ***** ***** ***********

** *** *******, ** ***** **** if * ****** ** ********** ** a *** **** *** *** ********* the *****, ** *** ****** *** classified ***** ** *** ******* ** all *******, ******** *** ***** *********** accuracy ****** *** ******* ** ~**-**%.

IPVM Image

Black *** **** ******** ******* ****** *** ***** ***********

**** ***** *** **** ******** **** the ******* ***** *** *** *******' color ***********. **** ***** ******** **** recognized ********* ~**-**% ** *** ****.

IPVM Image

Motorola ***** ******* ***** ** ******* ***** ******** ** *****

** * ***** ** ******** ** both *** *** *****, *** ******** camera **** ******* *** ******* ***** with *** ********** ***************, ****** *** detection ** ******** ** ** ***** be ** *** ******* ******* ** completely ******* *** *******'* **********. *** only ***** ** **** *** ******* plate *** ** ** **** ** the ******* ********* ***** *** ** reliable *************** ******* ***** ** ** visible ******* ** *****.

IPVM Image

Verkada ***** ******** ** ******

** *** *******, *** ******* ******* attribute ********* ****** ************ ********** ******. For *******, ** *** **** ********* to ****** *** **** ******** ** you ****** ** **** ***** ********, and ********* *** ***** ******** ***** only **** ******. **** ******** ********* with *** *** ******.

IPVM Image

Most ****** ***** ****, ****, *****: ****/******** ******* ** *********

*** **** ****** ***** *************** **** work/delivery ******** ** **********. ***** **** typically ********** ** ** *** ** truck ** **** *** ********** ** all, **** ** ******* **** **** a "**** *****" **************. ***** **** also *** **** ***** ************* ** missed *********** ** ******* **** ****/***** recognition, **** ** *******, ********, ** Rekor.

IPVM Image

Verkada ****** ****** ******

*** ******* ****** ******* *** * search ******* **** ******** **** **** eight ******. ****, *******, ********, *** Rekor ****** *** * ****-**-**** ******* to ****** **********, ****, *************, ***., or * ****** *** **** ******** many ******* ** ****** **** *** search.

IPVM Image

*******'* ****** ******** *** ******* ****** of *** ****** ***, ***** *** be ********* *** **** ** **** for ******* **** ** ******* **** is ***** ********* ****** ** *******.

IPVM Image

Analytic ***** **** **** * ****** *** *** *******

*** ****** ******* **** ******** ******* on ******* ********* *** ******* *** inclusion **** ***** ******** **** **** 8 ******.

IPVM Image

Vehicle ********* ****** **** ********* *** **** *****, ****

** *** *** ******* ******, ******** Alta *** **** **** *** **** ones ** ***** ****** *** ******** with ******** ********** **** **** ******* the *****. ******** **** *** **** offer **** ************* **** ** ***** is *********, ***** ** *** ******** and ******* *** ** ******* ****** in *** *******.

IPVM Image

Genetec, ********, *****, ******* **** ******, ** *****

*****, *******, *** ******** *** ************ for *** *** **** ***** ***** notifications *** ******* ***** *******, ***** Verkada **** ****** ****** *** ****** of ******** *** *** *******. ******* attributes *** **** ** **** *** search *** *** ******.

IPVM Image

Axis ****** ********* **** ****** **** ********** **** ***

**** ****** ********* ** *** **** vehicle ********* ********* ** **** **** that ********** **** *****. ***** **** do *** ******* ****** **** *****, they **** *** ****** *** ******** vehicles.

IPVM Image

Firmware ****

  • ******** ****:*.*.*
  • **** ***-**-*** ******: **.*
  • ******* ****** ***********: ** ******** *****
  • ******** ***: *.*.*
  • ***** **** ***:*.*.*
  • *******: ** ******** *****
Comments (4)
KH
Kyle Hoertsch
Feb 01, 2024

** ** **** ** ******* ******* and ******** ******** *** ******* ** that ****** *****. ******** **** ******** user ******* **** ******* **** * day ** *******. ***** ********* ** the ********* *** ***** *** **** a *** ** ********* ***** ** cameras ****.

(1)
UM
Undisclosed Manufacturer #1
Feb 02, 2024

***** *****, ****** *** **** ****. There's ****** **** ************* ** **** than ***** ** ****, *** ***** be ***** ** *** ****** ** evaluate******* **** **.

(3)
UI
Undisclosed Integrator #2
Feb 06, 2024

* ********** ****** ** *** *** the ****** ***-******* **** ******** ** the ****

UE
Undisclosed End User #3
Feb 12, 2024

* ***** **** ** *** *-***'* vehicle ************** ******** ********, ** **** as ******'* ******* ** ****** ***.