Startup Spectral Edge RGB+IR Claims "Unique Low Light Performance Tech"

By Ethan Ace, Published Apr 03, 2019, 08:34am EDT

This UK startup, born out of university research, is boldly proclaiming that they deliver "Unique low light performance tech for video surveillance systems", using machine learning.

Low light is certainly an issue for video surveillance, though while improving, still challenged to deliver color.

Now, Spectral Edge has introduced RGB+IR Fusion, aiming to use color and IR in the same image, delivering color in areas light is available, and monochrome IR in dark spots.

spectral edge~1

Inside, based on a conversation with their CEO, we examine Spectral Edge, their Fusion technology, how it compares to Hikvision's Color/IR DarkfighterX, and the company's future outlook.

Spectral **** ***+** ******

******** ****'* ***+** ****** technology ****** ******* ** simultaneously ******* ***** (***) and ******** ***** ** the **** ***** ** determining ******* *** ****** on *** ****** ****** be **** ** ******* color ****** **** ***** is ********** ** ****** to ***** ** ****** when ***** ** ***.

**** ******* ** **** time, **** ***** ******** from ***** ** ********** IR ** ***** ******, shown ****:

**** *** ******* *** advantages:

  • ***** **/*:*** ******* ********* ** that ******* **** ** "color ****" **/* ******* of ********* **** ***** during *** *** ** monochrome/IR ** *****. **** eliminates *** **** ** adjust ***/***** ********** *** lower ***** ***** ** when ******* *** *** angle ** *** *** change.
  • ** ********** *** ******:************, ******* ***** ******** Edge ** *** ******* an ** *** ******, eliminating * ********* ***** of *******, ********** ** lower **** *******, ** well ** *** ******* time ** ****** ******* video ***** *** ****** engages/disengages.
  • ***** ******* *****:*******, ******* *** *********** provide ****** ******** **** those ***** *** ****, since **** *** ******* additional ** ***** ******* affecting ***** *******, ****** exposing ****** ***** ** the *****.

Non-Standard ******* ******** (***-**)

*** ******* ** **** RGB+IR ****** ** *** compatible **** ******** **  sensors, *** ******* ******** image ******* ***** *** separate ****** *** *** and **, ****** ***** as ***-** (** ******* to ******** ***, **** known *******). *******, ***** ******* are *** ******, ***********************, *** ** *** largest ****** *********, ********** RGB-Ir ** **** * few **** ****** ** the *****-*** *****, *** none ** ****** ***********.

******* *** ******** **** time, ** ***-** ** relatively *** (********** ** the **** * *****), but *** ***, ** 2019, **** *** **** limited.

Spectral **** *******: ********* ** ******** *** *************

******** ****'* ** ********* this ********** ** *** manufacturers, **** ** ********* or *********, ** ******* to ************* ***** *** chips ** **** ***** own ******* ***** ******.

** **** ********* ****** is ********* *** ** not ******** ** ***** to **** **********.

*** ******* **** **** to ************** ****** ** impress ****** ************* ** that **** *** ** to ***** *** ********* and ********* **** ** add *******.

*******

*** **** **** ****** final ****** ******* ** unclear. ******* ****, ******** Edge **** **** *** run ** ******** *** chips. *******, ******** **** will ****** ********* *** their ** *** ****** manufacturers **** **** ** use **** ****** ***-** sensors.

** *** **** ****, it ********* ****** ** less ********* **** ***** two ******* *** *** sensors.

Compared ** ************

******** ***********'* ************ (******), ***** **** *** separate ******* (*** *** RGB *** *** *** IR), ******** ****'* **** structural ********* ** **** it ******** **** * single ******, ******* ** two ******, *********** *******. Because ** ****, *** potential *** *** ** lower **** ******* ** greater **** ************, ***** remains * ***** ** Hikvision's ****-*** ******* **** (~$2,000+).

******, ***** ******** **** produces * ****** ***** instead ** ********* ** and ***** ******, ** is *********** **** ****** to ****** **** *** haze ** **********-******* ******** seen ************** **** *******, ***** ******* ******* somewhat.

************, ***** ***+** ****** still ******** "******" ***** data **** *** ******, Spectral **** *** ******* better ***** ********, ** less ********** **** ** performed ******** ** ********* images ** ************.

Compared ** ***** ***** ************

******** ** ***** ***** illuminator *******, **** ** the*****/***** ****** ********** ***************** *******, ******** ****'* **** advantage ** *** *** of "*********" ******** ************, instead ** *** ***** white ***** **** ** these *******, ***** *** be * ******** ** users ***/** ***** *********.

How ** ******** ** ****** *******

**** ** ****** *** most ********* ******** *** the ******* ** ****, at **** *****. *** many ****** ********** *** premiums *** ******** *** light, *** ******** **** compares ** *****'* *** low ***** ********** ** hard ** *** ***** production ******** **** ***** cameras ** *** *****.

** ***** ******** **** for * ************* ** show ***** ********** **** by **** **** * conventional ****** *** **** supplied ****:

**** ** ********* ********** better *** ** ****** be ******* ** ** an ********** ********** ** finished ********** ********.

***** ** * ***** SpectralEdge ******** ***** **** clip *** ***** ****:

*******

** ** **** *** us ** **** *** well **** **** **** until ** *** ** with ** ** ****** cameras ** *** ******** from ****** ************* **** are ********** ****, ***.

*********, ********, ****** ** for *** ************ ** be ******* *** ********** as ** ******* *** range ** ******* *** potential *********** *** ******. 

Vote / ****

Comments (4)

Spectral Edge's is licensing this technology to SoC manufacturers, such as HiSilicon or Ambarella, as opposed to manufacturing their own chips or even their own cameras using Fusion.

Do they truly have blocking patents that they can license? Clearly, they did not invent the RGB+IR sensors, so their technology is primarily signal processing related? 

The demo of the partial IR+white light illumination is neat, but it feels like that would mostly be an edge case. Illumination would typically be similar across the entire image, it might change frequently, but having an image that is all RGB or all IR exposed for any given frame is more likely to be the case than frames that are half and half.

This feels like something between a patent troll and a lazy developer to me.

Agree
Disagree
Informative
Unhelpful
Funny: 1

This would be extremely useful in outdoor environments, if not required, once this gets a foothold in mainstream production. How many suspect vehicles can you eliminate when you've determined it was a Maroon Full-size truck vs a "dark-colored" truck? This feature could be a fantastic addition to the consumer market in something like a Ring Floodlight cam, where the transition from IR to Color is painstakingly slow. Same with any install outdoors where the light could change quickly, from floodlights, headlights

Not every camera in every application would benefit, but I could see this being a big perk for outdoor installs. An evenly lit scene is ideal and obviously makes life easier, but how often is that reality? 

Agree: 2
Disagree
Informative: 1
Unhelpful
Funny

I think the overall concept of RGB+IR video is very useful. I question the need to do individual images that are half and half, particularly to the extent that is causes a manufacturer to want to license some idea where the manufacturer still needs to do most of the work of implementation.

Agree
Disagree
Informative
Unhelpful
Funny

How many suspect vehicles can you eliminate when you've determined it was a Maroon Full-size truck vs a "dark-colored" truck?

Mark, good comment. I had not thought of that. I could see this being of use. Also, in areas where part of the FoV is illuminated by a streetlight and part is not is another example of that.

Agree: 1
Disagree
Informative
Unhelpful
Funny
Read this IPVM report for free.

This article is part of IPVM's 6,938 reports, 926 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