Yet another cheap thermal imager incoming.. Seek Thermal

[Fraser]

@Uho,

Very nice work. I applaud what you have achieved with simple, off the shelf lens types. Your lens housings also look neat. Are they made on a lathe, printed or constructed by some other means ?

From what you have said they invert so they are a Keplerian design. I have commercial lenses that are also Keplerian and I will either need to introduce erector lens groups or correct the captured images in post production.

If you are going to sell your designs you may wish to design adapter brackets to clip them onto the new FLIR ONE as it will be another healthy market for you. If the lenses could include an erector element pair they would be more marketable to the general public. I was going to use a pair of 12.5mm or 25mm FL Plano convex ZnSe lenses for the task. The down side, as you know, will be the extra physical length of the lens. The only way I know to shorten the length is to use all Germanium lens elements due to their large refractive index.

As I said..... Excellent work....... Respect

Aurora

[Uho]

Thank You. I tried to make nice. I made on a lathe. Lens X1.7 not turns. I used a negative lens. It is expensive and rare. It is difficult to buy. And it does not allow a large increase. The first lens has to be very large diameter. Why did only one lens for tests.

[frenky]

Do any of you have a good optics simulation software?
I was trying to simulate Seek's 3.8mm lens with additional 50mm ZnSe...
My ultimate goal is to find a configuration to convert Seek's non focus camera to a focusable one.
Could you help me with that? I have one 50mm and one 100mm ZnSe lens.

The only free (non java) software that I have found is this:
https://chrome.google.com/webstore/detail/ray-optics-simulation/egamlemiidmmmcccadndbjjihkcfiobh?utm_source=chrome-app-launcher-info-dialog

This is simulation of 50mm ZnSe lens 30mm away from Seek's lens:

[miguelvp]

When I did look at something that I needed for optics a long time ago I found out about OSLO

https://en.wikipedia.org/wiki/Optics_Software_for_Layout_and_Optimization#Free_Educational_Product

http://www.lambdares.com/oslo

Edit: Link to the OSLO-EDU version:
http://www.lambdares.com/oslo-edu

They had a free version, it was hard to learn and I forgot how to use it after I was done with it.

But I bet there are other free alternatives now.
Edit: and other easier alternatives for that matter.

[frenky]

Tnx. I'll check it out.

[lod]

I have unraveled the fix that Seek put in to correct for the edge gradient problem.

Essentially they use any shot of a flat surface to calibrate a correction matrix which is applied to all subsequent images.

This is done by taking blocks of sixteen pixels and calculating the average value.
These average values are again grouped into blocks of sixteen and the range between the highest and lowest value determined. This range is used to determine if the image is a flat surface to be a basis for updating the calibration.
The blocks of average values are again averaged, to get a relatively crude map of the image.

To create the correction matrix a full image is created by interpolating out the crude double averaged values. 7/8ths of the correction matrix is the previous value, 1/8th is the new interpolated values. This low pass filter prevents any glitches from having too much impact.

The corrected image is created by subtracting the correction matrix from the original.

To properly document this I have created an Octave/Matlab script. The script is not an exact reproduction of Seek's code but shows the algorithm which was used.

To get data into Octave I adjusted Seek_2.0.py by Cynfab to export a Matlab data file, I couldn't figure out how to work with the generated PNG images.

Source code, including invocation instructions, are on github:
https://github.com/lod/seek-thermal-documentation/tree/master/scripts/thermal_gradient

Floor - Before correction	Floor - After correction
Face - Before correction	Face - After correction

[Fraser]

Excellent investigation work.

Thank you for finding the answer to this outstanding question.

Aurora

[in.Vitro]

Hi everyone

at first: sorry for my bad english


Today UPS deliverd my Seek (Non XR-Version). Actually I want to buy the FLIR One (second gen.) but its very hard to order here in Germany... *sigh* I gived up to waiting 

And thats why I orderd the Seek for iOS. (FYI, I ordered the XR-Version 3 Weeks ago but sell it again and orderd the non XR Version to days ago!)

But today there was a little surprise: The non XR-Version has a focus ring!  I'm not sure if Amazon send me a wrong product.... I dont think so! Or it is a new Version. Because the Ring is much smaller!


Its hard to tell if the Focal length is similar to the XR-Version.. maybe you guys can help me

[frenky]

Thanks for sharing. 


This is definitely a new model:

[in.Vitro]

But the question is: Why is it labeld as "non-XR"? I mean, its obviously the XR because when I strech my arm and make a "selfie", my whole face filled the frame (yes, I have short arms )



The Ring on the "new" Model is about 5mm thick.. I would gues on the XR is the thickness about ~7,5mm? right?

[frenky]

Take a look at writings on the module.
Perhaps there is EAN or model name somewere on the module?

[frenky]

And take a look at the box to. EAN is on it for sure.

[in.Vitro]

On the Modul is just the Serial Nummer, CE and this annoying trash bin.


the EAN is  8 55753 00518 1

[in.Vitro]

Good News! 

Its a new Model (Hint: look at Amazon )

this little sucker has impressive macro capability!!
I hold the Cam ~3cm above the PCB

[frenky]

Looking good.

Photo from amazon listing:
http://www.amazon.com/Seek-UW-AAA-Thermal-Imaging-Connector/dp/B00NYWAHHM/ref=sr_1_1

[jadew]

Just got a new phone (an S4 mini) and it turns out it's not compatible with it. Any ideas on why that may be? It's very frustrating.

[frenky]

Devices with USB Host support:
http://www.gsmarena.com/results.php3?sFreeText=host

[jadew]

That's a bummer. I find it hard to believe that whatever SoC they are using in this phone, doesn't have support for USB host. Is that really the case or are they crippling them on purpose?

[frenky]

I'm not sure you should compare it to other phones with the same SoC.

P.S:
Another list of phones with USB Host:
http://goo.gl/47fjHo


EDIT: I did a quick search and it seems that none of the phones with "Qualcomm MSM8930AB Snapdragon 400" have USB Host support.
http://goo.gl/SmBdQS

[jadew]

I'm not sure you should compare it to other phones with the same SoC.

P.S:
Another list of phones with USB Host:
http://goo.gl/47fjHo

Good idea, thanks.

Edit:
A quick look over some threads at xda-developers.com would suggest that they crippled them on purpose. The hardware supports it, but they're not providing any power from the phone, so you would 1) need firmware and 2) have to provide power to to the device from an external source.

[frenky]

Yeah I was researching in that direction too for my old LG One P500 and HTC Desire.
But in the end I decided that is to much trouble because there is plenty sub 150€ phones that support it out of the box.

Sub 150€ phones:
http://goo.gl/avdcSs

[encryptededdy]

Yeah, the same issue exists with the Nexus 4 - data but no power.

I guess that's why the new FLIR One has its own battery.

[minim]

I have read through 26 pages of this thread and I will keep reading but I just wanted to check if anyone might have tried to do the same as I want to do with this sensor and get a analoge output.

I have been helping my neighbour locating his animals on two occasions where people forgot to close the fence and once when someone sabotaged the fence and let his animals out in the mountain. I do this by flying with a RC-plane rigged with a video link and flying over a large area looking for them. I've seen that a SAR group here in Norway had a drone equipped with a FLIR camera and when searching for people or a animal herd they seem to find them really easy due to this expensive thermal camera ($4-5k). So I wanted to install the same system on my drone and help local farmers and possibly the local SAR group with search missions. I'm funding this on my own and it's not for the purpose of making money so I can't really afford to pay $5-10k for a camera. This $250 camera and some gear for making analoge video out would be perfect. I guess the purpose of my mission isn't on-topic but the analoge video out part might be a little bit on topic

Edit: is raspberry pi and using the audio port for analoge video out an option or isn't it possible to use the PI to read video from the sensor?

[Fraser]

Minim,

The distance capability of the SEEK camera is quite limited. The Microbolometer Array passes its analogue pixel levels to a Read Out IC (ROIC). This ROIC is part of the Microbolometer core. The output from the ROIC is not analogue on the SEEK.

The only way to get an analogue signal from the SEEK for your needs is to uses a small dedicated computer such as a Raspberry Pi to translate the image to composite format.

Drones such as you have described normally use a FLIR TAU or similar. There is good reason why they need such expensive cameras.

Aurora

[ez24]

I have read through 26 pages of this thread and I will keep reading but I just wanted to check if anyone might have tried to do the same as I want to do with this sensor and get a analoge output.

I have been ...  on-topic but the analoge video out part might be a little bit on topic

Edit: is raspberry pi and using the audio port for analoge video out an option or isn't it possible to use the PI to read video from the sensor?

I suggest you start a new post and re-word your question.  This post is much too long to make any sense.  I think you will get more answers if you put more of your question in the title.  If you do somehow let me know.  The animal rescue is interesting.  Also set your country flag via your profile, members appreciate this.  I suggest this title :  "Can you use a Pi drone with a Seek Camera for remote animal rescue?"  Then paste your project into the post.  Good luck