Author Topic: The SEEK Thermal core development - user opinion please  (Read 3296 times)

0 Members and 1 Guest are viewing this topic.

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
The SEEK Thermal core development - user opinion please
« on: February 04, 2020, 01:31:37 pm »
I was one of the first to purchase the SEEK Thermal phone dongle type camera severall years ago.

I was somewhat excited by the prospect of a competitor to the FLIR lepton core that could introduce healthy competition and choice to the marketplace. I have always hoped that thermal imaging equipment would become available to the masses and far more affordable to organisations like the Search and a Rescue services. The SEEK camera appeared to offer much and at a very affordable price.

Once I received the First Generation SEEK camera dongle I was impressed with its compact dimensions and metal case. It looked like a lot of thought had gone into the product and I hoped it was SEEK dipping their toe into the market with a view to developing market share and a positive reputation on which to build. Bearing in mind the very knowledgeable team heading the company (one being the founder of AMBER and Indigo) it looked a sensible business plan for a startup company. Start small and build a product range over time.

Testing of the SEEK dongle was relatively painless with the application and camera working well with my less than stellar Motorola MOTO G generation 1 phone. All seemed to be working as it should but there were some issues that reduced my enthusiasm for the camera. Remember, I am used to using thermal cameras that cost upwards of $35K so my expectations of the little SEEK Thermal may have been a tad unrealistic considering the price point for relatively high resolution. The FLIR E4 that I purchased before the SEEK camera had impressed me as I had upgraded its lowly 80 x 60 pixels resolution to full 320 x 240 pixels and switched off the artificial noise injection. That sadly influenced my view of the SEEK camera and, on reflection, that was unfair of me. The E4 was a sub £1K camera but was only 80 x 60 pixel resolution. Yes I upgraded it to the E8 specification, but that was a far more expensive camera from FLIR. The SEEK offered higher resolution than the E4 (as bought) and at a significantly lower price. It was $200 if I remember correctly. That is quite some achievement by SEEK.

So what were the issues that I disliked about the SEEK Thermal dongle camera ?

My first impression of the images produced by the camera was that it was very ‘laggy’ and there was a lot of random noise present. The lag may well have been partly due to my mobile phone not being exactly a fast model ! The noise was a very real presence but in retrospect I think I was being a little unfair as my brain was used to viewing nice low noise images coming from vastly more expensive cameras. Once the camera was used to image a scene with greater thermal Delta T, such as a car engine or cup of coffee, the noise level reduced as the temperature span was automatically increased by the app. The images were then pretty good considering the cost of the unit and its tiny size. There was still some noise in the images but the camera was performing well enough for many users needs.

There was a proven technical issue with the camera that caused a thermal gradient across the image. There was some investigation of the matter on this forum but I do not think the cause was ever totally proven. It could have been FFC shutter related, thermal creep on the PCB, or something else. SEEK Thermal responded to the issue by providing an updated APP and, as if by magic, the gradient disappeared from the images produced. SEEK never released information on what they had done to remove this problem but the pragmatist in me says, provided the camera behaves as expected and within specification, does it really matter what they did ?

I sold that SEEK Thermal dongle camera to someone who needed it more than me.

Since my brief ownership of that SEEK Thermal product I have sadly been a little biased against SEEK products as I already own over 80 various models of thermal camera and remember the noise on the original dongle was a little distracting to me. Remember, it was a Gen 1 core in that unit.

So move forwards a few years and I start reading comments on improved imaging from SEEK Thermal products. The PRO version of their core was released and this has higher resolution (320 x 240) with reputedly lower noise levels. Images I have seen published from the PRO cores cameras certainly seem better than I remember coming from the first Gen 1 SEEK dongle camera. I once made a somewhat negative comment about the SEEK Cores to a good friend who designs thermal imaging cameras for a living. He immediately commented that I should look at the images produced by these more recent SEEK products before discounting their performance. I was shocked to hear him say this, but pleased in equal measure as I have always wanted SEEK to succeed. It is the perfectionist in me that is to blame for my jaded opinion, formed by experience with a very early version of the SEEK core. Well just before Christmas I met up with my thermal camera designing friend and he placed a SEEK Reveal in my hand to try :) The SEEK Reveal product has several advantages over the phone dongle in my eyes. It is personal preference, but I like a thermal camera to be a complete self contained solution that I can rely upon without the worry of phone related issues or unwanted app updates that change the UI. The Reveal is a SEEK Core that is integrated with a dedicated microprocessor that focuses on creating a decent thermal image with the specified frame rate. The performance can be better tuned in such a design than may easily be achieved in a dongle that uses generic configuration to work with a vast array of Android mobile Phones. Note I say Android because iOS phones are effectively a standard for which a dongle may be tuned for best performance (FLIR advised me of this when developing their F1G2) The Reveal is always ready for action, reasonably robust and the screen size is acceptable for the intended use. The relatively small screen, compared to a mobile phone, helps to improve the apparent image quality as well ! Then there is the LED torch. I thought this a total gimmick, but I was wrong. On a number of occasions I have wished for an integrated torch when viewing a target in the thermal domain, but low visible light. That torch could be a godsend to many users. Not a gimmick in my opinion. The FLIR Exx series effectively has a torch as well :)

So there I was, holding the SEEK Reveal with its sub QVGA resolution and I actually liked it ! My friend was right, the SEEK Reveal Thermal camera was actually a lot better than The early dongle version I had used. Sure there was still noise on the image when viewing a scene with low delta T but when viewing a ‘normal’ scene with warm targets present, the image improved and less noise was apparent. Was it as good as a multi thousand Dollar thermal camera from FLIR or one of the other big names in the Industry ? Well no, but that could be considered a very unfair comparison ! If you are rich enough to buy a new FLIR E8 or E75, you are unlikely to be looking at the SEEK Reveal anyway ! The SEEK Reveal is also compact, easily carried. I actually went home from my meeting with the SEEK Reveal with a completely different view of the more recent SEEK Thermal products. I have yet to use a SEEK Thermal PRO based camera but hope to do so in the near future. I expect to be pleasantly surprised by the Price to Performance ratio :)

What potential buyers of SEEK Thermal core based products need to consider is what they need the camera for and the price being asked by SEEK. If the price to performance ratio is good, then it is a sensible purchase. There is some choice in the market of phone dongle and handheld thermal cameras these days. The price of a Therm App or Thermal Expert may be too high for some people or they may want an integrated solution rather than a Dongle + Phone + App that is the Therm App and Thermal Expert topology. With time we will hopefully see even more choice in the market and I am seeing an increase in Chinese ‘home grown’ cameras coming to market.
This can only be good in the long term for the consumer as prices will fall. With the release of the SEEK Thermal cores for 3rd party use, we are likely to see more manufacturers using that relatively affordable core in all manner if equipments and camera designs. FLIR Lepton seems to have made slow progress in this respect so I hope SEEK Thermal does better.

So after that Loooooong introduction, I would like to hear from users of SEEK Thermal products and their experiences with them. Is the PRO core as good as I am hoping and if not, where does it fail to satisfy you. Is the SEEK dongle improved over the original Gen 1 ? If so, in what ways ?

Basically I would love to hear your thoughts on the SEEK Thermal products and cores. I promise to be open minded and I do now realise that I may have been unfair in my previous opinion of the cores. Tell me how it really is please  :-+

Thanks

Fraser
« Last Edit: February 04, 2020, 03:00:00 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 
The following users thanked this post: nikitasius

Offline Ultrapurple

  • Super Contributor
  • ***
  • Posts: 1027
  • Country: gb
  • Just zis guy, you know?
    • Therm-App Users on Flickr
Re: The SEEK Thermal core development - user opinion please
« Reply #1 on: February 04, 2020, 03:57:48 pm »
I was as underwhelmed as you with the early dongles, for the same reason - relatively poor IQ. Noise, relatively poor resolution and poor response to small temperature variations (ie poor contrast in challenging conditions) were my main issues. In early 2016 I did a side-by-side comparison of a Seek XR that I'd bought used and a ThermApp camera with different lenses. The primary reason was to compare the field of view but it became clear that the Seek IQ left something to be desired. (Click here for the original posting on Flickr, which includes a description of what I did, plus access to the full size image).



The Seek XR was capable of producing reasonable resolution images using stitching techniques, as seen below - the red box indicates the actual sensor FoV. But that technique is only any good for non-moving subjects and doesn't address the sensitivity and noise issues, though if you've got time you can address the latter using averaging of multiple exposures. (If the image below had been made as a 3x3 grid of original images then a total of about 300 exposures would have the potential to produce about a quarter of the noise and possibly a slight increase in resolution - but if you took one second per exposure, you're looking at five minutes for image capture plus a significant amount of faffing around to de-noise and tile the exposures to get a noticeably better result).



According to the published figures, the later Seek sensors are lower noise (effectively = more sensitive) and, as noted by Fraser, higher resolution, so I would expect the resultant images to be substantially better than those I was able to achieve with my gen 1 XR.

Like Fraser, I do hope so, because it's high time we saw some decent quality, low(ish) cost thermal imaging devices on the market. (Sidestepping but acknowledging the well-known Problems with The Bad Guys).

Whilst I do like my ThermApp dongle-style cameras (I keep one with me almost everywhere I go, unlike my much bigger & heavier FLIR SC-600), I can also see the attraction of the all-in-one does-what-it-says units. 20 or so years ago Garmin revolutionised the GPS market with its eTrex series of hand-held mapping GPS devices, well before today's Google Maps/Earth etc. In about 2001 I bought my first eTrex, for about $99, in a Chicago Wal-Mart and immediately used it to navigate the 900-odd miles from there to Portsmouth, Virginia (and back). I can see the Seek devices - which are a similar form-factor - doing the same thing for thermal imaging.

At least, that's what I'm hoping. 
« Last Edit: February 04, 2020, 04:13:04 pm by Ultrapurple »
Rubber bands bridge the gap between WD40 and duct tape.
 

Offline Bill W

  • Super Contributor
  • ***
  • Posts: 1108
  • Country: gb
    • Fire TICS
Re: The SEEK Thermal core development - user opinion please
« Reply #2 on: February 04, 2020, 04:17:20 pm »
The latest SEEK cores are in some 3rd party cameras now, 3M-Scott V320 series.  Much of the data is however geoblocked from the UK, so just the YouTube sales videos





and there is little detailed TIC footage as it does not record or do video out.

Note that Scott X series are the last of the ISG parentage cameras (so use a ULIS sensor) and the Scott Sight is a FLIR LEPTON.

Bill

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
Re: The SEEK Thermal core development - user opinion please
« Reply #3 on: February 04, 2020, 04:53:09 pm »
Hi Bill,

I found a pretty cheesy 3MScott booth video that contains comment on the V320.

https://m.facebook.com/3MScott/videos/introducing-the-3m-scott-v320-thermal-imager/2288211311412974/

From what I could see, the frame rate looked decent in that camera. I am aware that the Reveal XR30 sold exclusively in the UK was capable of 30fps which was impressive when it is normally around 15fps for seek products. As you say, no really detailed reviews or pictures of the V320 image quality but still an interesting development. I am aware of one fire department that equipped all 24 of its fire fighters with the Fire fighter orange version of the SEEK Reveal. The FLIR K1 costs around $600. With prices of thermal cameras becoming more affordable as core prices drop, it can only be a positive future for personal thermal cameras. Where used in situations involving safety there will, of course, be a need for minimum acceptable performance levels.

I have high hopes for the SEEK PRO cores that are being made available to 3rd party manufacturers. The FLIR LEPTON is, of course, another option, but its use is under the control of FLIR’s marketing and business strategy teams so possibly not so easily used as the SEEK products. FLIR will want to protect its own products from serious competition.

May the 12um microbolometer development continue at a pace  :-+

Fraser
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
Re: The SEEK Thermal core development - user opinion please
« Reply #4 on: February 04, 2020, 04:53:46 pm »
Ultrapurple,

Thanks for the very interesting comments and pictures etc  :-+

Fraser
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
Re: The SEEK Thermal core development - user opinion please
« Reply #5 on: February 04, 2020, 05:12:37 pm »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
Re: The SEEK Thermal core development - user opinion please
« Reply #6 on: February 04, 2020, 05:19:24 pm »
Comment from a FLIR agent on lower cost thermal imaging cameras for use by those helping the public......

https://www.firerescue1.com/fire-products/thermal-imaging/articles/why-thermal-imaging-cameras-have-become-essential-tools-n4x03cVfuw2SVZJ2/

Fraser
« Last Edit: February 04, 2020, 05:22:09 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline IwuzBornanerd

  • Frequent Contributor
  • **
  • Posts: 319
  • Country: us
Re: The SEEK Thermal core development - user opinion please
« Reply #7 on: February 04, 2020, 08:57:07 pm »

Well, you have seen my report on the improvement made between 2015 & 2017 Seek XR:
    https://www.eevblog.com/forum/thermal-imaging/latest-seek-compact-(non-pro)-has-significantly-reduced-noise/

  Since that report I also bought a Fast Frame Pro at the end of 2017 and a Slow Frame Pro back in September of 2019.  From what I can tell in the units I own the noise level is the same on all of the units since 2017.  I believe I have posted comparisons with the FF Pro on the forum but I don't offhand recall where.  All such comparisons were made using my own software since until recently I had nothing to run the Seek app.  But the latest version of Androidx86 I now have on this laptop will run the Seek app fairly well.  Since I like to back what I say with photographic evidence I have downloaded the latest (January) version of the app & will try to post some photos off the Seek app comparing the 4 units using the same low contrast scene.  I might not be able to get exact span-span comparisons because the Seek app does not allow the temperature span to be set for the non-pro units.  It could be next week before this happens given my current situation, though.

As for the comment about the amount of thought that went into the Seek dongles, I agree.  We have seen elsewhere on the forum that owners of other brands would appreciate having a bulkhead-mounted USB connector like on the Seek.  I think also that the bayonet mount Seek uses for the lens might be intended to facilitate a future move to interchangeable lenses.  The current configuration does not allow a typical owner to make such a swap but it seems like it would not take much of a change to make that possible.
I am not opposed to exercise, unless it is an exercise in futility.
 
The following users thanked this post: Fraser

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
Re: The SEEK Thermal core development - user opinion please
« Reply #8 on: February 04, 2020, 09:20:59 pm »
IwuzBornanetd,

Many thanks for reminding me about your excellent tests. That was what I remembered and what gave me hope that SEEK had managed to improve their imaging cores performance. It is interesting that performance remains unchanged since 2017 but if SEEK are satisfied with the current noise levels they may be sticking with their 2017 design of microbolometer and noise processing.

Fraser
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline Bill W

  • Super Contributor
  • ***
  • Posts: 1108
  • Country: gb
    • Fire TICS
Re: The SEEK Thermal core development - user opinion please
« Reply #9 on: February 05, 2020, 12:40:22 pm »
The FLIR K1 costs around $600.

For those less familiar with it, the K1 uses a 160x120 Lepton with MSX and extended temperature range to 400°C

Yes, that's MSX in the dark, in smoke and for not very hot fires.

Bill

Offline FraserTopic starter

  • Super Contributor
  • ***
  • Posts: 13263
  • Country: gb
Re: The SEEK Thermal core development - user opinion please
« Reply #10 on: February 05, 2020, 12:59:11 pm »
Bill W,

Totally agree, the FLIR K1 is not what I would call “Fire rated”  ;D I do like the fact that the big manufacturers are looking at budget priced fire/SAR cameras though. As previously stated, any camera involved in safety work really should meet an agreed minimum standard of performance and durability. For SAR activities on open ground an affordable thermal scope with decent performance would be a godsend to the search teams. If the SEEK cores can achieve the required minimum performance for such tasks, it will be good news. They already offer the Seek Reveal Shield model for police and security use.

Maybe we will see cheaper cores coming out of China in the years to come ? The SEEK cores are still a US produced product and that remains a thorny issue for some manufacturers and distribution. It would be wonderful if SOFRADIR-ULIS produced a competitor to the LEPTON and SEEK Thermal cores as that would improve the current situation immensely.

Fraser
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline IwuzBornanerd

  • Frequent Contributor
  • **
  • Posts: 319
  • Country: us
Re: The SEEK Thermal core development - user opinion please
« Reply #11 on: February 12, 2020, 11:18:21 pm »
I guess everyone else is waiting for my samples before posting their own comments, huh?

So I'd better get on with it. :)  I'm having trouble deciding what I need to say here so I'll keep it short & let questions come as they may.

I present 2 sets of images.  Interested persons should look at both.  The first set is from the Seek app running under Androidx86 on my laptop computer & the other set is from my own software running under Debian Linux on the same laptop. 

EDIT: NO the system swapped them on me!  And after posting 4 times already I'm not going to try fixing it!  The first set is off My software.

These images were taken shortly after warmup and a fresh NUC (2-point NUC that is, or "gain" correction).  The individual images in each set are labelled as to which camera they are from.  The Seekware images are taken with version 2.1.6.3 of the software.
 
I managed to find a sufficiently low contrast scene that would cause the Seekware to set a temperature span of 3 or 4 degrees Centigrade, which is kind of my standard for comparison.  Unfortunately the Seekware does not distinguish fractions of a degree so if it says the max is 25 & the min is 21 for a difference of 4, is it really 4 or is the max actually 24.55 & the min 21.45 for a difference closer to 3?  And if an image from another camera hits the other sides of the half mark the difference could be close to 5 degrees.  I'm not sure that is enough to account for the difference between the images from the 2 pro cameras, though.  But then with my software the Slow Frame Pro appears just a bit noisier than the FF Pro but with a wider 4.6 degree span vs. the FF Pro at 3.5 degrees.  Hmmm...

The stupid system keeps screwing up with the attachments so I'll try again...and again with them at the end as in the old days.
« Last Edit: February 19, 2020, 01:26:46 am by IwuzBornanerd »
I am not opposed to exercise, unless it is an exercise in futility.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf