Genetec Cloud LPR Examined

Published Apr 10, 2015 04:00 AM

The cloud is yet to be a significant factor in video surveillance.

Now, Genetec is trying a different approach, trying to simplify LPR using the cloud.

In this note, we examine the features, limitations, pricing and competitive positioning of Genetec's new cloud LPR offering.

**********

******* *** *** *** *** *** cameras *** ******** *** **** ***** (see******* ****** *** ******). *******, ** *** **** ******* to *******'* *** *******. ****** **** support **** ****** *** ******** *********, they *** *** **** ********* **** conventional ** ******* (~$*,*** *** *******'* LPR ******). ************, ******* ******* ~$*,*** for *** ****** *********.

Cloud ***

***, ******* ** ****** ******* *** conventional ** *******. *** ***** **** be ******** **** *** ******* ** site, ********* ** *** ***** *** results **** **** ** *** ***** Genetec ******** ****** ******.

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

  • ******* ***** ** ** *** *********. Higher ****** ***** ******* ***** *******'* ****** LPR *******.
  • ****** ****** **** ** ****** *** approved ** *******. ******* **** ******* ***** *** *** **** *****-** are ******* ******** *** ***** ****** from **** *** ********* *** ******* soon.
  • **** *** ************ ******, ***** **** / *** ***** *********** **** ** is ****** ** **** ******, ** caution ***** *******.
  • ***** *** ** ************ ******* **** not ******* ******** ********* ********* ** Genetec ****** ***, **** ******* *****, direction ** ******, *** ******* **** identification.
  • ** ****** *** ***** *** ********** for ************ ******* *******.

*******

******* **** **** ************ * ******* ************ *** ** USD $**-$** *** ******.

Pros *** ****

*** ******* ********** *** **** **** eliminates ******* ** *** ********** ** the ****** ****, ***** *** **** some **** ***, **** ***********, * fair ****** ** ****** ****. **** a ****** ******* ***********, ** ********** the ******** ~$*,*** *** ****** *********.

** *** ******** ****, ** ** limited ** ******* ***** *** * few ************ *******. ~$*** *** **** is *** ***** *** ***** *******'* high *********** *** ********* ***** (**** annual ***), ** ***** **** *******, it ***** **** *** ** ** **** expensive.

*******, ***** *****, ********* *******'* ******* rival ********* (************ *** ******), ******* ******* ***** ***** *******, allowing *** ***** **********, ****** **** a ********** ***** *** *** **** to ****** **** ********* *** *** overloaded.

*******, ** *** **** ******** ** Genetec's ****** ******* ** ***** *** for ************ *******.

UPDATE ******

** ********** *******'* ***** *** ****.

Comments (6)
SM
Steve Mitchell
Apr 10, 2015

You say "the video will be streamed from the cameras on site." I kind of doubt this is what's happening. If so, they'd need to identify the bandwidth requirements.

The ideal architecture here is to use object recognition on camera with a classifier that isolates the license plate (or at least indentifies candidate frames that include a vehicle). Then send images to the cloud for recognition--most likely using a machine learning system.

Seems like that type of implementation would lead to the narrow camera requirements listed.

Maybe I'm giving Genetec too much benefit of the doubt, but that's a good architecture as it allows many tags to be identified most efficiently, and the cloud-side recognition can be tuned and modified without need for on-site intervention.

(1)
U
Undisclosed #1
Apr 10, 2015
IPVMU Certified

Using either software triggers, such as motion or cross-line detection, or hardware triggers, including in-ground loop and IR sensors, AutoVu Plate Reader Cloud will detect vehicles and upload a short encrypted video feed to its cloud-based engine, where the image analysis is performed. - Genetec

Supported video stream resolution 800 x 600

Note: Though most advanced analytics are not supported, vehicle make identification is.

(3)
Avatar
Jean Pierre Picard
Apr 13, 2015

Hi Steve,

John has properly described the way Plate Reader Cloud will function. The local Security Center system streams a short video archive when a trigger detects a vehicle. The entirety of the image analysis is performed in the cloud, including license plate detection. This limits local processing requirements to its minimum and enables seamless updates to the license plate-finding component of our ALPR engine.

Plate Reader Cloud was designed to address one of the core inefficiencies of software-based ALPR solutions, which is that you must always have enough processing power for peak activity times – typically the beginning and end of office hours. During the rest of the day, this (costly) processing capacity is simply unused. By moving image analysis to the cloud, Plate Reader Cloud eliminates the need for off-peak processing overcapacity, and decreases the total cost of ownership of a software-based ALPR solution.

We feel this approach will both let customers scale their systems without having to invest in significant local processing capacity and achieve the high accuracy and performance they expect from Genetec AutoVu solutions.

Jean-Pierre Picard
Product Marketing Manager, Genetec

SM
Steve Mitchell
Apr 13, 2015

Jean-Pierre, thanks for the clarification. I read John's original description as all video may be streamed continuously to the cloud recognition services. This is obviously not the case, as an on-camera or other trigger is used to first identify candidate frames, then imagary is uploaded to the cloud for recognition. You're saying, rather than individual frames, it is a short series of frames. Which I interpret to be basically the same architecture I'm describing. :)

Sorry to split hairs--I tend to distinguish between a contiuous stream of video to the cloud vs a file being uploaded to the cloud. They have very different implications in terms of bandwidth and reliability.

Can you say more about the narrow supported camera requirements? I assume this is due to the need for object recognition or tripwire capabilities built into the Q1765 and P1425-LE? Is there any camera software licensing required?

Avatar
Jean Pierre Picard
Apr 13, 2015

Hi Steve,

You are right, video is not streamed continuously, and the final architecture is very close to what you described.

The list of cameras certified at launch will go beyond the Axis Q1765 and P1425-LE, and will include cameras from both Panasonic and Sony. While the selection might seem limited at first, we are confident the final list will provide enough options for customers to meet requirements in a wide range of scenarios. Additional brands and models might also be added post-launch.

Customers associate Genetec’s AutoVu product line with high accuracy. In our experience, consistently attaining a high level of accuracy at day and night is only possible with a restricted number of cameras, and requires testing that goes beyond validating a camera can output a stream at a pre-defined resolution and framerate.

Certification is 100% focused on real-world ALPR performance – each camera on our list has been installed at our own offices for several weeks before being certified, and the specific camera settings required to achieve high accuracy will be documented in the Plate Reader Cloud installation guide. Object recognition or tripwire analytics are not part of the testing process.

As for licensing, no additional camera software license is required beyond the subscription fee mentionned in the original article.

Thanks,

Jean-Pierre Picard
Product Marketing Manager, Genetec

BT
Brian Tally
Apr 14, 2015

I see this being huge in city wide deployments. Being able to strategically place LPR throughout a city without the huge upfront cost of Sharp cameras will draw more cities into making investments in LPR.