Lepton 3.5 "loosing definition" over time

[Andrew Jones]

Hi,

I've had my FLIR Lepton for just under a week, and during that time I've started to notice that my Lepton has begun to "lose definition" over time. What I mean is that when I first received the Lepton, the images captured had quite well definite "background scenery"; however, over the last 24/36 hours or so, I've noticed a significant drop in the definition of my images. So much so, that I've gone from having a background with clearly defined objects, to just having a pretty much white image.

I should add, my Lepton has been running pretty much non-stop attached to a PureThermal 2 over the last 5 days or so; there have probably been some times when it has been off for > 12+ hours, but in the last 48 hours it probably hasn't been off for much longer than an 30 minutes ("maintenance" of the software I'm creating). Furthermore, and just to give context, I am capturing images at around the rate of 1 every 20 seconds.

Has anyone else experienced this? Is this to be expected, or do I hav

e a potential hardware issue?

I tried to find some images that illustrate the problem, and I've attached an older one (with definition) and a newer one (loss of definition) to this post. Regrettably, I mostly delete images after 24 hours, so the "really clear" ones are gone. As you can see, the items are more well-defined in the older image, and completely washed-out in the newer one.

Cheers,

Andrew

[Paul S]

Hi Andrew - do you have a shutter on your Lepton? If so you need to ensure its working and perform a NUC (Non Uniformity Correction). If you don't have a shutter, then place a sheet of uniform material with high emissivity (matt painted piece of metal works well) in from to the lens and perform a NUC. This will return the image to the original sharpness. (Alternatively switch off the LEPTON, place the sheet in from and switch on again, once the image is stable, remove the sheet and your scene should be sharp again. The LEPTON should have a scene based NUC but continuous use like this is probably causing the drift - a useful explanation of the problems associated with Spatial noise is given here: https://www.flir.com/globalassets/imported-assets/document/quark_ssn_description.pdf - its not specific to the lepton but the physics is similar.

[Andrew V. Jones]

Hi Paul,

Interesting. As far as I’m aware my Lepton _does_ have the shutter,
because it is just an out-of-the-box Lepton 3.5 without any
modifications.

Is there anyway I can check the status of NUC with a Lepton that has a
shutter?

Cheers,

Andrew

> --
> You received this message because you are subscribed to the Google
> Groups "Flir Lepton" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to flir-lepton...@googlegroups.com.
> Visit this group at https://groups.google.com/group/flir-lepton.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/flir-lepton/6668db29-a2a5-469c-b25a-3b49a1cab0d9%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Paul Sacker]

Hi Andrew -

Lepton with Shutter: This will have a black (or silver) rectangular plate on the front of the module, with a hole in the middle showing the lens material, clipped on via some lift off tabs on the edge - if its radiometric it should have the shutter


Lepton w/o shutter: this will have a direct view of the lens so the front of the module will have a round element with the lens in it.

rgds

Paul

[Andrew V. Jones]

Hi Paul,

Yep! I checked — my 3.5 is indeed a radiometric one and has a
plastic-looking black plate on the front.

How can I check if NUC is indeed “working correctly” on my Lepton?

Cheers,

Andrew

[Andrew V. Jones]

Hi Paul,

Ah, good idea about GetThermal — I actually hadn’t even thought to
check the imagery with that!!! I will give that a go :)

Cheers,

Andrew


On 10 May 2018, at 11:10, Paul Sacker wrote:

> Hi Andrew - when you turn it on you should see the shutter “wink”,
> and it should then periodically “wink” if you have it enabled. If
> you use GetThermal there should be a button to enable the FFC (Flat
> Field Correction) - I’m coming from a Thermography background so
> writing code is a black art I have yet to master, but using cameras
> etc I’ve been doing for years. I’m now trying to work out how to
> use a Lepton 2.5 for a device to capture a 14bit image in a scene with
> a span of approx 10-15 degrees….its proving tricky for an old
> dog….
>
>> On 10 May 2018, at 11:05, Andrew V. Jones <andrewv...@gmail.com>

[Andrew V. Jones]

Hi everyone,

Does anyone know how frequently I should see the shutter actuate? I’m
trying to test if it is firing at all on my Lepton.

Thanks!

Andrew

[Andrew Jones]

Hi again,

I tried playing with PureThermal to resolve this issue, but I didn't have much success. Indeed, performing FFC inside of PureThermal seems to have no affect on the device at all!

Am I correcting in thinking that hitting the FFC button inside of PureThermal should actuate the shutter? Should this actuation be visible to the human eye?

Cheers,

Andrew

On Thursday, May 10, 2018 at 7:32:19 AM UTC+1, Andrew Jones wrote:

[Kurt Kiefer]

That's correct, you should see and hear the shutter when you press the FFC button, it's pretty obvious.

From the datasheet, here's when the Lepton ought to be activating the shutter:

• At start-up
  
• After a specified period of time (default of 3 minutes) has elapsed since the last FFC
  
• If the camera temperature has changed by more than a specified value (default of 3 Celsius degrees) since the last FFC
  

Even if the shutter isn't working, you should be able to put a uniformly heated surface in front of the Lepton and hit the FFC button to run the correction. This is what you do with non-shuttered 50deg and 25deg Leptons.

[Rod Christel]

Hi everyone,

Kurt is exactly right on the shutter timing.  When the Lepton is powered up you should see (and usually hear) the shutter cycle within the first few seconds of operation.  The shutter will close for a little less than one second.  Normally you can see the glassy round lens in the center opening in the front of the camera.  When the shutter cycles a flat black plate will slide over the lens.  Powering up is the easiest time to see the shutter activate automatically, so you don't have to try not to blink for several minutes.  Pressing the FFC button on the PureThermal GUI should also trigger, but there are a few other possible causes of problems in that case.

Shutter problems usually create a different image problem though.  If the shutter sticks open (fails to cycle) at first you see a completely blank image.  If the camera is then moved, you see two images superimposed.  One normal, moving image, and a second, frozen, "negative" image of the scene that was present when the FFC occurred.  This combination is a perfect diagnostic indication of a shutter problem.   If the shutter sticks in the closed position of course you see just a black image regardless of camera motion.  In that case you should be able to see the flat, black shutter blade instead of the lens.  If the shutter is not activating all the way, it is possible to have a low contrast image something like the image above.  If it is sticking half way you should be able to see the blade of the shutter any time you look.  If it is just not closing all the way, you should see some "ghost" imaging when the camera is moved as when the shutter is stuck open.

It is difficult to see exactly what is going on without knowing what the scene and setup are, but it seems that one possible explanation is a change in the weather.  If outside, sunshine or cold nights would make large differences in temperature in a hutch, while overcast or fog can cause low contrast images because there is less difference in the temperatures of things in the scene.  Even if indoors, the slight breeze from a heating or air conditioning system will make for better contrast.

For full disclosure, I should say that I work for FLIR, supporting OEM users of thermal camera modules.  Note that I am not an official spokesperson.  We have taken an interest in this issue, and will be contacting Andrew off-line to help him resolve this issue.

Best regards,

Rod