Ranking Video Analytic Architectures

Author: IPVM Team, Published on May 13, 2016

Video analytics may be implemented in a number of fundamentally different ways. Many ask: Which one is best? Each has important tradeoffs. Moreover, in our experience, some are far more likely to be accurate as well as simpler to deploy.

Below, we rank and analyze five fundamental approaches (alphabetically):

  • Add-On camera analytics
  • Analytics appliance / encoder
  • Cloud analytics
  • Embedded camera analytics
  • Embedded DVR / NVR / VMS analytics
  • Server based analytics

We also look at recent advancements in video analytics and what may be available in the future.

Let's start with a review of each approach:

***** ********* *** ** *********** ** * ****** ** ************* different ****. **** ***: ***** *** ** ****? **** *** important *********. ********, ** *** **********, **** *** *** **** likely ** ** ******** ** **** ** ******* ** ******.

*****, ** **** *** ******* **** *********** ********** (**************):

  • ***-** ****** *********
  • ********* ********* / *******
  • ***** *********
  • ******** ****** *********
  • ******** *** / *** / *** *********
  • ****** ***** *********

** **** **** ** ****** ************ ** ***** ********* *** what *** ** ********* ** *** ******.

***'* ***** **** * ****** ** **** ********:

[***************]

Add-On ****** *********

**** ** **? ********* ******** ** ****** **** ** ** ******. ***** ** analyzed ****** *** ****** *** ******/****** *** **** ** *** systems *** ******* / **********.

**********

  • *********** ** **** *** ****** ***** ******* *** ***** ** run ********* **.
  • ******* ** ****** ********* ****** ************ ****** **** ***** ****** into * ******** ****** ******.
  • **-****** ********** ******* ****** ******** ************

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

  • **** * ******* ** ************* ********* ******* ****, **** ******* Axis.  *****, *** ********* ****** * ******* *** **** ********* by **** *******.
  • ********* *** ********* *** ********* *** *** ******** ****** *****.
  • **** ****** **** *** ******** ****** *** ** ********** **** one's ***

****: ****** ***** **** ***** ********* ******* ***** * *******. We ********. ***** ** *** **** *********, **** ** ********* not * *** *******. *** ******* ** ****** *** ***** stream *** ** ***** ******* ****** ** ********* *** ******* using ***** ********* **** *** **** * *** **********.

**** ********* *** ******** ** ****?***** **, ***** **********.

Analytics *********/*******

**** ** **? ******** ******** ** ****** ***** ********* *** ******** ****** *********.  A ********* ******** ****** ******* ******** ******** ******** ** ***** stream **** * ******. 

**********

  • ********* ******* ******** ********* *** ***** ***** **** ******-***** *********.
  • ****** *****/*********** *** ** ******** ******.
  • *** *** ********-***** ********* ** ******* **** ** *** **** the *********/******* ** *** **-***** *********.

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

  • ****** **** ******** ** ***-** ** ******** *********, *** ** being ********* ********.
  • ***** **** ******* *** ************.
  • *** ** **** ********* ** ************ *** ** *** ** camera ****** *** ********* ******.

**** ********* *** ******** ** ****?******** (******),*** (******)

Cloud *********

**** ** **? ***** ** **** ** * ***** ******* ******** *** ********.

**********

  • ****** *** ** ********** ****, **** ******* * '*******' ***** stream ***.
  • ***** ******** ***** *********** ** ******* ***** ******* ****** ** add ************ ** ******** **-****.
  • ***** ********* *** ***** **** **** ** ***** **** ** them ** ******* ***** ********* ***********.

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

  • ******** ********* ** **** ***** ******** ** *** *****, ***** is ***** ******* / *********, ********** ** **** ******* **** a **** *** ***** ****
  • *** ** *********, ********* ******* ******* ****
  • ******* ****** ** ********* *** ******** ****

** *** ****** (******* ****** ****), **** ** ****** ** become * **** ****** ******, **** **** *********.

**** ********* *** ******** ** ****?******* ***** ***,*****,**** ***

Embedded ****** *********

**** ** **? *** ************ ******** *** ************* * ****** **** ***** *** embedded ***** *********.

**********

  • ***********: ************ *** ******** *********** ** ******** ****** ********, ******** and ******.
  • ********** ** *****: ** **** ** ******* ******** ******** ** new ********.
  • ********** ** ************/*******: ***** * ****** ****** ** ****** ****, easier ** *** ***** ******* ***** **** ** *****.
  • ******** ** *** ******* ** ******** *** *******.
  • **-****** ********** ******* ****** ******** ************

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

  • ******* ** * ******* ** ********* **** ***** ****.
  • ****** **-*** ******** '*******' *******.
  • * *** ***** ************* **** ***** ******** ****** ********* "*********" which ** ********** *** ****** ** *******.

*** ** **** ****** ********* **** *********?

****** *********-**** ******* **** **** **** *** ********** *** *****, generally * ******/**** *** * *********** *********.  **** ******** ****** detection ******* *** ** ** *** ** ************ * ****** tripline-style *****.  

**** ********* **** **** ********-***** *****, **** ** *********, ********* of ******, ** ****** ******* ****** ******.  **** ********* **** also ****** *** *********** ****** **** ****** ** **** ** the *****, ************* **** * ******* **** ****** **** ** smaller **** * ****** ***, *** ****** *********** ***********.  **** true ********* ******* **** **** ****** *******, ******* **** ***** can ** ******* **** ** ******, ** **** ** ********.

**** ********* *** ******** ** ****? *** **** ***** ********* ******** ** ******* - ********, ****, Bosch, ****/*******/*****, **********.

*** ****** ********* ******** ** ***** ********* - **** *** Samsung ***** ******.

Embedded ***/*** *********

**** ** **? ********* ******** ** ***-****** ** * ***/*** ********* ** ****** VMS ********. ************** ****** ******* ***/** **** ** ********* *** their ******* *******.

**********

  • ****** *****: ** ******* ** *****, ** ********** ******** ** install *** ** **** ** ********* **** ******** ********
  • ********** *******: ******* *** ********* *** **** ** *** ******** solution, ********* *** ********'* ****** ******** ****** ********** ******, ****** and/or ********* ******.

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

  • ******** ******: ******* *** ******** ******** **** *** **** *** quality ** *************** ** *** ******* *********, ** ****** ******** its *********** *** ****** ******** ******.
  • ***** **** ** *******: ***** ********** ******** ** **** **** H.264 ******* ***** ************* ******* ****** ** ***** ******* ****.
  • *********** ***********: ********** ** *** ** ***/*** **********, *** ********* may **** *********** ************ ** ***** *********** ** ******* *********.
  • ********* *** ******: *********, ***** **** ***** ** ********* *** numerous ******* *** ******* **** * ******** ** *****. ********* to ****** **** ** *** *********.

**** ********* *** ******** ** ****? *******, ***.

Server ***** *********

**** ** **?********* ******** ** ****** ** * **/****** **** ** ******** from *** ********/*** ***** ****. ***** ** ******** ** ** the ****** *** ******/****** *** ********* ** *** ********'* ****** or **** **** ** * *** ** **** ******.

**********

  • ****** **** ********* ** *** ****** ** *** ********* ********* inside ******* *** ********* (***** ** ***** *******).
  • ********** *** ****** *** **** ** ** ******** ****** ********** of *** ******* ** ********* ******* ** *********.

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

  • ******** ****** ******* ***** ** ********, ********* ******** ******* ** many ******* **** *** *********
  • ********* ******* ********* - *** ****** ***** ** ********** ****** from *** ******
  • ******: ******* **** ******** * *** ****** *** *** ******** is *****, **** ***** ** ** *** **** ****** ********.
  • ******** ******: ******* *** ******** ******** **** *** **** *** quality ** *************** ** *** ******* *********, ** ****** ******** performance *** ****** ******** ******.
  • **********: ******* * *** ****** ***** ** ** *** *** integration **** * *** ** **** ***** ** ** *****, this *** ** ***** **** ********* ** *** **.
  • *********** ******* *******: ********** ******/*********** *** ****-**** ****** ***** *** not ** ********* *** *** *** *********

**** ********* *** ******** ** ****?***** ****, ***********, ***, ********

Rating ***** ******** *************

*** ***** **** ***** ******** ****** ********* ** ***** *** most ******** ******* ***********, **** ********* ********** ******** ******* *********** when **** **** ******* *******.

** **** ***** ********* ************* ** *** ********* *****, ******* in **** **** ** * **** ** ***** *** *** absolute:

  1. ******** ********* - ******* *****, ***** ****, ********* ****** *********** *** beware ****** *** ********** ** ** *********.
  2. ********* ********** - **** *********** ** *******, *** ****** *****.
  3. ***-** ********* - ********* ****** ***********, *** ******* ******* *******. 
  4. ***/*** ********* - ****** *****, ******** ********** **** ***. ***** lower ***********.
  5. ******-***** ********* - ******** ****** ***********. ****** ** ** ******* cost, ******* ***** *** *** ******* *********** ***** **** ***.

****: ***** *** ******* **** **** *******, ** ** ** too ***** *** ******* ** ********* *** *** **** ** bad ** **.

***** **** ******* ***** ******* ******, ***** *** ***** ** video ********* *****, *** *** ******** *** **** ************ ** ******** performance. *** ****** **** *** ******* **** ***** ********* *****. A ******* **** ***** ** ****** *** ******** **** **** do *** **** **** ****** *** ********** ***.

**** ********** *******, ***** (** *************) **** ***** ***** ********* ** maximum *****.  ***** ***** ** *********, ***** ******** **** ****** also ** ********* (**: ** ******** ****** **** ******** ***' *****, but **** **** ** **' **** ********** ***** ** **** may ** **** ****** **** ** ******** ****** **** * 300' *****, *** * ***' **** ***** ** ****).  ***** should **** ******** ******* ********* ********** ***** ** ***** ***********, in **** ***** ** ** ********* ** ******* >***' ***** consistently ** *** **** ** ***** ** ****, ****, ***, dust ******, ** ***** *******.

************, ***** ***** ********* ****** ** **********. ** ********* * 500' ********* ***** ***** ** *** **** ** * **** number ** ***** ******, *** ****** *** *** ** ** unreliable *******.

Recent ******* *** ************

***** ********* *** ******** ************ **** **** ********** ****** *********** over *** **** ******* *****.

**** ******* ***** ******* ***** ** ****-****-********* **********, *** *** few ******* **** ****** ****** ********** ****** *** *** ****** 2MP (****: ***** **** ********* ******* *** ******** **** ** 4K ** ****** ***********, **** ****** **** ****** *** *****, this ** ******** ** *** **** **** ****** ******** **** does *** ******** ***** * ***** **********).

****** ********* ******* ****** ********** *******, ******* ***** ******** *** ****** as "******" *** **** ********** ******* ******* ******* ("***/***** *******", "truck/large *******").  ***** ***** ** ****** *******, **** ** "****", "bicycle", "********", "******", "********", "***", ***. *** ****.

************ **** ********* **** ** *** **** ** **** ** deployment *** *************, **** ********* **** ******* *** *****/*** ********* auto-calibration ** ****** ***** **********.

****** ******** *** ********, **** ******* ****** ****-**** ******* ********** power **** ******, ***** **** **** ******** ** ***-** ********* even **** *********.

**** *** **** * *****, ** ****** ************ *** *********** ** ***** ********* *** ******** ************ ** improve *************. *** *******, ******** **** ****** ******** ***** ****** *** ********* ***** ********** *** ******-******* **********.  ****-******* ******* ************ are **** ********* ******* ******** ** ****-**** ***** ********** *** object **************. ***, **** *** *** *****, ***** **** ******* with **** **-***** ********** ***** **** **** ***** ********* **** viable.

[****: **** **** *** ********** ******* ** **** *** ************* re-written ** **** ** ******* ******** ************ ********** ** ****** side ********* *** ***** *********.]

Comments (16)

***** ********* ** * **** ******** **** ** *** ***** of *****'* *********. **** * **** **** ** **** * know ***** **** * ***** ***.

* ***** ** **** ********** ** *** *** ******* ** the ****** ****- *******. *** *********** ***** **** *** *** been ****** ** ******** ** * *** ** ** ******* it ***** ** ** ***. * *** **** ****** ** is *** ** *** *** ******* ** *** ******** *** putting **** ***** ** *** *&* ******, *** ** *** same **** ********** *******. ** ** ***** **** ** * lot ****** **** ** *****!

******,

*** **** ******, ** ***** ** *** ******** *****, *** for ***** *********. ** ******* * *** ** **** *** report ******* *** **** ******** *** *** *********** ******** ******* and ******** **** *** ******* **** *********.

* ***** **** *** **** ** *** **** * *********** of ******* **** **** **** *** ********* ****** ****. ****** seem ** ** ** *** ******* *** ****** ******** **** a **** ********* ****.

***** ********* **** ** ********** *********. ******** ******* ******* **** at ***** ******* ************** *** **** ** ****** ***** ********* and **** ******* *** ************ ** *** ****/********. ********* **** locally *** *** ********* **** (***-*******) ****** ******** ******* * warrant ** *** ******* ************. **** ***** ********* *** ** important ** ********** ******** ****** ****** ** ************. ******* ** not **** ****** ** * ********** ************* *** **** **********, must ** ********** ** *** ********* **********.

** *** ***** ******** ******** **** *** "****** *****" **, due ** *** ******* ***** ****** **** *** *****? ** if **** ** ****** *** *** ** ** ** ******, cheaper, *** ********* ******** ** *** *****...*** **** *** *********** due ** "******* **************"? ****'* ****.

******** ****** ** *** "****-******** ***** *********" ******** ** *** the ****-******** (** ****, ** *****) *** *** ******** *** demonstrative ****** ******* ** ********* ***** *********. *** ******* ***** positives **** "******* ** *** ***********". ** ** *** **** - ** **** ** **** ** **** *** ******/ ********* to ****** *** ****** ********. ********* ******* *****'* **** ** know *** ** *** **. ** *************/ **** ***** *********. Is **** **** ********* ****?

**** ************* ** *** ********'* ********* *****/****.

** ****** ****, *** ****-******** ** ********* *** **********. ** only ****** ***** ** ********* ***** ********. ******* ******** *****'* do ******** ** ****** **** ********, *** ** *** ***** is ****** **** *** ******** ******* **** **** ****** **** a **** *****.

*** *****-**-******* ******* ** ******* **** *** ***** ** ******** and ****** ******** *** *********** ** *** **********. ** * "good" *********** *** *******'* **** ** ** *** *****-**-******* ******* of ******.

*** ****** *** ***'* **** ********* ********/**** **** *** ****** to ******* ** ******* ***** ***** ** *********. *** **** about *** ******-************** ********* ******, *** ****** ** ********** ** establish ** ************* ** ****** ****, ***** *** ******* **, mounting ******/*****, *** **** ***. **** ** ** **** ** can ******* * *** ** ************ ***** ***/***** ******* ****** appear ** *** ***** ** ****, *** ******* *** *****/***** objects ****** ** ** ********* *****.

** *** ****** *********** **** **** *** ****** ** *** next *** **** *** **** **** **** ***'* **** ****** worse ******* *** ******** ** *** ***** ***** ** ***** different.

** ********* ** *** ***** **** ***** ** ******* ******* mounted ****** *** ********* ** * **** ** ***** *********. If *** ******* ******* *** *** *** **** *****, ****'** have *** **** ***** ******, *** ** *** ******** ****** and **** ****** *** ****** ***** *** ***** *********** *** by **** ******* *********** **** **** *** **** ** *** next.

*** ******** ********* ************, ** *** **** ****** ***** *** are ******* * *** ** ***** ****** *** ******* ** often ****** **** ** * **** ** **** ********** ***** activity ** *** *****. ** ***'** ***** * ****** (** or ******), *** ******* *** **** ** **** ** **** input *****, ******* **** ***** *** ** ***** ** ***** problems ** *** ***** ****** (**** ********, *** *********) **** impair *** ********* **** "******" **** ** ****** *****.

******,

*******, ** ***** ** ***** ********* ** ********, ********* *** banners - **** **,*** ******* ** *** ********, **** ** the ******* **** ******** ****** *** ****? (****'* * *** a ********* ***** **** ** *** **** ***?). **** *** been ** ***** ** ***** *** ********** **** ***** ******, encoder **** ** *******/ ******** ** *** **** * *****. The ******** *** ** ***** ******** ** **. ** ** the ****** ******* ****** (***** *** ****)** ***** ********* ** our ********** *******.

* ** * ****** ********* **** *** ******** *** ** real ***** ****** ** **** ********* ** *** **** *****. Can *** ********* *** ******* ** *** ******** **** ***** improving *** *********?

*, ****'* * **** ********. *'** ********* ** ** ******** for *******. *** * ***** **** ******* **** **** **** as **** *** **** ** ***** *******?

***** ********* ***** **** ****** ***** *** ******* / ******** analytics ***, *** ******* ******* *********** ******* (**** ******** ********, etc.), ** ***** ** **** *********** *** ******** *** ******** to ******* ********.

****,

* **** ****** ** ********** *** ***** ********* *********** ***** to ***** ****** * ***** ** *** ********** ******* **** the ***** *** *****. ** *** **** * ** ****** drive **** ***** ******* ** ********, *.*.

******'* ** ********** ********* ** ***-** ****** ********* ** **** you *** ********* *** ********** ******** ****** ******** *******? ** my ********** *** *** **** ************* ** **** ****** *****.

***, **** ** * **** *****, *'** ****** *** ****** to **** **** *****.

****** **** **** *** *********** ** ********** ** ******** **** article. ** * **** ** *** ******** ****, ***** ** no ***** *********** **** *** ************* ******* ** ******* ********* into * ***** ************* ****.

* ** ******* ** *** **** ** ********* ** *** http://avss2016.org/ **********. * **** ** ***** ****** **** ***** ** deep ******** *******.

** **/**** ** ****** ** **** ***** **/** ** ********** 820 ** ***** ** ***'* *** * ****** ** *** solution ***** ** *** ********** ****** *******. ** **** ** server **** ********* ** ** **** ******** ** ****** *** something **** ***** *** *****-******* ********** ** *** ******** ****** GPU ****** (**** *** **** ** ***** ** ***** ** or **** ******** ******). **** *** **** ** **** *** to ******** **** *** ** ******** **** ** ******* **** their ****** ******* **** ** **** *** ****** ** *** reasonably ****** ****.

***** **** **** * *** ********* ** *********://***.***********.***/

** **** ********** ** ********* *************, ** **** *** **** omitted *** ****** ** ********* * *** **** *** ********** analytics. ** **** *********** ** **** ****** ***** ********* **, I ****** *** *** ********* * *** *** **** *********** a *** ***** ********* ******** ** ** **** **. * agree **** **** ***** ******* **** *** ***** ** **** its *** ** *** ** ***** ***** *** ** *** cost, *** **** * *** **** *** *** *** *********, you ***** ****** **** **** *************. * *** **** ****** something, *** ****'* **** * *** **** **** *******....

  1. ******** ********* - ******* *****, ***** ****, ********* ****** *********** but ****** ****** *** ********** ** ** *********.
  2. ********* ********** - **** *********** ** *******, *** ****** *****.
  3. ***-** ********* - ********* ****** ***********, *** ******* ******* *******.
  4. DVR/VMS ********* - ****** *****, ******** ********** **** ***. ***** ***** ***********.
  5. ******-***** ********* - ******** ****** ***********. ****** ** ** ******* cost, ******* ***** *** *** ******* *********** ***** **** ***.
Login to read this IPVM report.
Why do I need to log in?
IPVM conducts unique testing and research funded by member's payments enabling us to offer the most independent, accurate and in-depth information.

Related Reports

Directory of 30+ LPR / ANPR Providers on Feb 21, 2018
License Plate Recognition / Automatic Number Plate Recognition are a type of video analytics software that can identify and match license / number...
Bosch Merges Video, Intrusion and Access Businesses on Feb 19, 2018
Bosch is merging their "video systems, intrusion detection, as well as its access control and management software business units to form a single...
Why 3VR Failed on Feb 16, 2018
3VR destroyed transformed ~$65 million in VC funding into a $6.9 million exit. The reason they failed is simple. They bet on analytics. They...
Last Chance February 2018 Camera Course on Feb 15, 2018
This is the last chance to get into the Winter camera course, starts next Tuesday. Register now. IPVM provides the best education, live online...
Hikvision DeepInMind Tested Terribly on Feb 15, 2018
While Hikvision is heavily marketing deep learning and 'AI' as their next big thing, new IPVM test results of their DeepInMind NVR shows their deep...
Assa's Lowest Power Draw Maglock: Securitron M680E Examined on Feb 14, 2018
Securitron produces some of the most extreme maglocks on the market, including massively strong maglocks and even ones with integrated CCTV cams...
Hanwha Wisenet X 5MP Camera Tested (XNV-8080R) on Feb 13, 2018
Wisenet X is Hanwha's high-end camera line. We tested their Wisenet X 1080p camera last year. Now Hanwha is offering 5MP cameras listing super low...
Favorite NVR / VMS Manufacturers 2018 on Feb 12, 2018
There is a new integrator favorite VMS. In 2016, 2 VMSes were effectively tied for top choice. One of those VMSes favorability was stable while...
Motorola Targets Chinese With Avigilon Acquisition on Feb 09, 2018
Motorola joins the growing list of companies taking aim at Chinese manufacturers. Recall, last week it was Hanwha: Chinese Products Damaged...
Arlo, Bigger Than Avigilon, More Valuable Than Axis on Feb 08, 2018
Arlo, the wireless IP camera offering that Netgear bought ~5 years ago for a few tens of millions is now doing more revenue than Avigilon and...

Most Recent Industry Reports

Aruba Networks Profile on Feb 22, 2018
Aruba Networks' presence in the video surveillance market has historically been limited. With a company focus on Wi-Fi first and switching...
US Army Base Specifies 70+ Outdated Hikvision Cameras on Feb 22, 2018
A US Army base has specified 70+ Hikvision IP cameras, a month after the WSJ reported a different Army base removed Hikvision IP cameras. While...
Directory of 30+ LPR / ANPR Providers on Feb 21, 2018
License Plate Recognition / Automatic Number Plate Recognition are a type of video analytics software that can identify and match license / number...
New Whole Foods Installs Hackable Access Control on Feb 21, 2018
Whole Foods has built a reputation for high quality. And their 2017 Amazon acquisition has increased that, plus added deep pockets for buying...
Remote Network Access for Video Surveillance Guide on Feb 21, 2018
Remotely accessing surveillance systems is key in 2018, with more and more users relying on mobile apps as their main way of operating the system....
Visio For Video Surveillance Design on Feb 20, 2018
Many integrators have standardized on AutoCAD for camera layouts but new users may be overwhelmed by its learning curve. Microsoft's Visio...
Health Care Insurance Integrator Benefits Statistics on Feb 20, 2018
How common and how much healthcare coverage is typically provided by security companies? 150+ integrators explained how their companies provide the...
Hikvision Deletes Genetec Support on Feb 20, 2018
There will be no peace between Hikvision and Genetec. A year after Genetec expelled Hikvision (and Huawei, citing Chinese government control...
Change Orders - Sometimes Necessary, Sometimes Unethical on Feb 19, 2018
Change orders are a common element in project sales. Sometimes they are a necessity and appropriate ways to deal with arising issues, but sometimes...
Bosch Merges Video, Intrusion and Access Businesses on Feb 19, 2018
Bosch is merging their "video systems, intrusion detection, as well as its access control and management software business units to form a single...

The world's leading video surveillance information source, IPVM provides the best reporting, testing and training for 10,000+ members globally. Dedicated to independent and objective information, we uniquely refuse any and all advertisements, sponsorship and consulting from manufacturers.

About | FAQ | Contact