Hanwha Face Mask Detection Tested

By Zach Segal, Published Jul 01, 2020, 09:58am EDT

Face mask detection or, more specifically lack-of-face-mask detection, is an expanding offering in the midst of coronavirus. Hanwha in partnership with Turkey's A2 has now released an option, but how well does it work? And how does it compare to Avigilon's offering?

IPVM Image

Inside this note, we share our findings across:

Executive *******

******'* **** ********* ********* worked ****, *******, ** suffers **** ******* ***** support *** ******** ********** high ***** ******* (~***** in *** *******) ******** usefulness. **********, *** *********, it ****** "*********** **** Mask *********" **** ** our *******, ********* ******* ~70% ** *** **** masks **** **** ***********.

Limited ****** *** $*** *********

*** ** ******** **** works ** * ****** models (~** *******) ***** older ******* * ***** and *** **** ******* from ****** ** $*** per ****** *** * perpetual *******.

Upcoming * ****** ******** **** *********

************, ****** **** * P ****** ****** (** the **+) **** ******* mask ********* ********* *** a **** ******** ******* later ** *** ******, specifically ***** ******:

IPVM Image

Comparison ** ********

******** *** ********** ** price (****), ***** ******* (much *******), *** ******** range (********* ****** **** than **** *** ***** density ** ******).

*** ********** ********* **** Avigilon's **** ********* ** that ** ****** "*********** Used **** *********". **** correctly ******* ~**% ** cases ** ********** **** masks ** *** *******.

***********, *** **** *****, we ** *** ****** people ** ****** ******** or *** *** ****** the ***** *** **** specific *******.

Comparison ** *******

******** *******, *********, *** ****** tablet **** *********, ****** *** **** more ******** ** ********* people *** ***** *****.

Performance *********

*******, *** **** ********* from ****** ** ********* well ** **** ** anticipate ** ** ****** use *****. ** ******* picked ** **** ********** in *** *** ****-*** settings, **** **:

IPVM Image

*******, *********** ******* ********** in * ***** *** scenario. **** ******* ** a *******'* ****, *** Hanwha ***-**** ****** ***** failed ** ****** **** violations ***, *** *** flag * ******** ******* walking ******* * **** room **** *******.

IPVM Image

** ********** ********* ** IPVM *******:

**** *** ** ******** with ******'* ******* ** fine-tune *** *** ********* ... *** ******** * better ***** ******* *** the **** **** ********* application

****** *** ** ******* included ********** **** **** detection. ** ***** **** when * ******* ****** by **** ***** **** exposed, ~**% ** *** time **** **** *******.

IPVM Image

** ********, *** ****** failed ** **** ** a ******* ******* ******** at *** **** **** it. *******, ** **** a ~*-****** ***** ** with * ***** **** tilt, **** ***** **** away. ***** **** ***** be ***********, *** ****** could ** ******* ** mitigate ****.

Will *** ****** **** **** ******

**** ********'* ********, ******* with ***** **** ****** will *** ***** * "no-mask" ***** (** ******* the ********* *** * human *** '***' * mask). *******, ** **** only ******** * ~**°***** of *** **** ** flag * **** *********.

IPVM Image

********

*** **** ********* **** Hanwha *** ** ********** was ****** ** ***** than ********'* ********. ** arm ******** *** **** could ***** *** ******.

IPVM Image

*******, ** ***'* ********** spoofing ** ** * major ******* *** **** consumers.

Monitoring **** **** *********

*** ****** *** ** monitored ******* ******* **** or *** *** *********** interface (***** *****).

IPVM Image

********** *** **** *** straight *******, **** ******** to *** ******* ******, options ** **** ******* events, *** *** ******* to ************* **** ***** for ******** ****** ***********, so **** ** ***** are ***** ******'* **** integrations.

PPF/Size ***** ******** *** ****** **** *********

*** ******* ***** **** ~70 *** *** ******** for **** *********. *******, a2 ********** ****** **** mask ********* ******** ******* 120px-160px **** *****. ***** we *** ******* **** the ****** ***-****, **** would **** ***** ***** only **** ** ********* distance ** ~*-*' *** mask *********, ****** **** the ~*'*" (~** ***) we ****** ** ** effective *****. ****** ***, the *** *********** ** significantly ****** **** ********'*, making ********'* ******** **** better *** ***** *****.

Comments (2)

*'* ********* ******* ** know **** *********** *** end-users *** ** *** use ******** (*** ****** is ***** ** **** are ** ******** ********) if **** ***** ****** to *** ** ** server ** *** **** detection ** ******** *******, or *** * ***-**** camera *** ********* ******* that *** ** ** the ****?

*** ***** ******** **** ties **** *** *** above ** ** **** end ***** ***** **** cover ****** *****, ***** one ** *** *********** in * ******** ** store, ** ***** **** want *** ***** ******* within * ******** ** have **** *********?

******* ** *** **** cameras *** ****** ***** the ** **. ** you ******* *** ****** it ***** ***** ** swap *** ****** *** move **** ****** ** a *** ******** **** does *** **** *** AI.

** *** *** ***** 100 ******* **** ** makes ***** ** **** up * ******.

Read this IPVM report for free.

This article is part of IPVM's 6,653 reports, 896 tests and is only available to members. To get a one-time preview of our work, enter your work email to access the full article.

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