Yet another cheap thermal imager incoming.. Seek Thermal

[MLutz]

here is the same pic as before with °F, sorry, i forgot that...

Yep, it will be fun to see what they did wrong in his new app if we'll have decompiled sources if possible of this new "sick" version.
Anyway emissivity settings are the same as in previous 1.4.0 version and correct for skin?
Did you tried install previous version and set the same emissivity, temp units and compare?
I can't believe they did not noticed something like this during testing 

BTW: md5sums are fine, so we have the same versions of "Sick" Thermal 1.6.0 apk 

lol @sick thermal... 

i checked emissivity and there is no difference!
when i uninstall 1.6 and reinstall 1.4.2 temps are back to normal...

btw. with 1.6 the gradient does look better as with 1.4.2, but not as much as i hoped!
but i'm with aurora: for this money this is a very nice device. i'm also into thermal imaging for years, and i did not expect to see a device in this price-range so soon...

regarding the temps on 1.6:
so there must be something strange going on, because i also think that this is something they SHOULD notice during testing!

can someone else beside Yushir0 report this problem as well?

edit: seems you are right that is has something to do with emissivity: now i'm again on 1.6 and there is NO MORE OPTION TO ADJUST EMISSIVITY!!!

how can they miss that (to have no option could be intentionally, because a change from skin to granite didn't change anything at 1.4.2 temperature wise) at QC??

[Mudd]

Yes,
Same problems ,temp way off,my coffee was about 86 degrees celsius and personally,I don't see an improvement in gradient issue..

[frenky]

I took my Seek to the optometrist (my ZnSe 4" Focal length lens showed up today).
And my seek can read the label in my router. Edit: (The red Z doesn't show neither does a big red V over the whole verizon logo) the letters showing are white/silver and they probably dissipate heat quicker.

Hi.

Could you please tell me the algorithm for getting this great images?
I'm in the process of making software for fine tuning the processed image.
I'm pretty sure the algorithm is not just: sensor - calibration + 16384 because the vertical lines are still very visible...

Tnx.

[eneuro]

now i'm again on 1.6 and there is NO MORE OPTION TO ADJUST EMISSIVITY!!!
..
how can they miss that (to have no option could be intentionally, because a change from skin to granite didn't change anything at 1.4.2 temperature wise) at QC??

It could be interesting to put on glass plate (table) a few objects with different emissivity like very low for aluminium foil , black plastic stuff, wood, etc, let it settle at constant ambient room temperature for longer time, measure this ambient temperature and than record this thermal objects scene quickly starting from lets say 2m distance to closer eg. up to 0.2m (20cm) at different angles around those things.
They should have room temperature if avoided reflections and human IR source, I guess, so looking what Seek app in old version 1.4.x and "latest" 1.6.0 will record could be very interesting.
BTW: Skin emissivity is about 0.99  according to this table: http://www.optotherm.com/emiss-table.htm
Polished aluminium 0.05-0.1 only, etc...

Edit: Changes made in new Seek apk release 1.6.0 in its libSeekware library used by java below:

Code: [Select]

$ diff libSeekware-1.4.0.2.so.nm.txt libSeekware-1.6.0.so.nm.txt
49a50,52
> aslpf_deinit
> aslpf_init
> aslpf_process
154a158,160
> gradient_deinit
> gradient_init
> gradient_process
249a256
> retrieve_bplut
266a274
> tDiodeCount
325d332
< _Z16median5x_processPtS_ii
As we can see added gradient_* functions and aslpf_* and  retrieve_bplut and a few other.
Versions v7a of libSeekware.so have the same functions exported but a few differences in internal math calcs etc, but exported symbols the same 

[frenky]

I have finally received my SeekThermal.

In the meantime I've made an app to fine tune raw file conversions. (App screenshot in the attachment)
So far I'm not near the results that Miguel is getting but it's a progress...

[heavybarrel]

Just an FYI if no one has seen it yet.

As you may know, we still cannot ship cameras to our Apple customers. We sincerely apologize for this and are diligently working on obtaining the proper certification so we can ship.

We are confident that we will be receiving approval soon, but in the meantime we will cancel all iOS orders and any credit card authorizations or charges. We will notify prior customers once we are ready to begin shipping and offer priority fulfillment, a 10% discount on the new order, as well as free shipping.

Again, thank you for your support.

Seek Thermal Team

[Yushir0]

Being that this new version is broken so severely, I would be willing to bet that someone at seek made the classic noob programmer blunder of building this apk off of some broken fork of the codebase that was being used for testing instead of the fork that they actually meant to push. If that is the case, maybe we'll get the "real" version sometime next week.  If that isn't what happened,  I don't know how they would push such an obviously unreviewed version to the market.

[eneuro]

@Yushir0 Hopefully, Seek firmware is not affected by this latest release, so we can develop our own open source version of thermal interface for Seek dongle  without any problems

[Mudd]

@eneuro

There was no mention of camera firmware upgrade during install of latest software.Should be good then.

[Fraser]

@eneuro,

You wondered about the possibility of X Ray imagery of the SEEK PCB. I quickly fired up the Faxitron, set it for maximum output (35KV) and grabbed some shots of various items including the SEEK.

From the X Rays it is obvious that the Faxitron can penetrate a standard PCB but it does get stopped by any steel plates etc. It shot through the Magnesium SEEK case without much trouble but the PCB appears to be quite dense, possibly with a copper thermal sink built in ? Faxitron X-Ray Imaging is frankly useless on the SEEK. It may improve if I removed the case but if you look at the RS232-Ethernet PCB images you will see significant difference in the image detail. 35KV should have made it through the SEEK PCB had it been a conventional build.

The down side of X Ray imaging at 35KV is that it provides the least contrast and detail at the detector array (camera) I can get better images and the RS232-Ethernet (without a number) image was taken yesterday at a lower KV and to my eye appears better ?

I have not used a low KV X-Ray machine before and I am still experimenting to get the most out of it. Sadly time is in short supply at present. The BIOPTIC camera software provides all manner of analysis and image enhancement options that I have yet to play with. Sadly if the X-Rays do not penetrate the DUT then no amount of enhancement will help. For info the BIOPTIC array Camera captures 2048x2048 pixels with a physical image capture area of 100mm x 100mm. Resolution capability is pretty good as a result.

X-Ray images attached for your interest.

Aurora

[Fraser]

Todays RS232-Ethernet images.

[eneuro]

@Aurora Wow, thx you found a time to make such great X-ray imaginery of this dongle 
Lets see not visible 


Edit: This Seek profile below is interesting-are there any rectangular holes in lens holder while it looks like there is something like this ?

This Seek dongle lens holder was disassembled and put together or it is oryginal untouched?   

@Mudd I've already checked MD5s of those two Seek MPU firmware files *.bin and are the same as in previous *.apk releases, while for the moment it is more important for me any changes in firmware than android app.

Lets monitor when new Seek patch to this broken 1.6.0 apk will be released 

[blackboxdisease]

Those rectangular holes at the lens is to let the shutter open and close I think.

[Fraser]

@eneuro,

Nice work on the images 

My SEEK is still a Virgin, unmolested and case still tightly sealed 

I can take some better pictures if needed from different angles. The second image was not great as the SEEK was just sat on its side at an angle to the detector array. There is a single rectangular shutter aperture in one side of the lens 'tube' and the X-Rays are punching through the other side wall producing the contrast where the aperture is located.

I took a look at the formats that the Faxitron software can save in and I am pleased to say that it offers "Lossless" JPEG, BMP, TIFF and RAW.

In the images you can see a possible reason why the Faxitron was taken out of service. One or two of the cameras panels are more grainy than the others. I will need to check that this is not something that can be "tuned out" via calibration settings. From my point of view it does not spoil the images so is no great concern.

Not on topic I know but some might still find the technology interesting  ....

If anyone is interested in what the Faxitrons/Bioptic camera Teledyne detector array looks like, take a look here:

http://photonics.engr.uga.edu/xray_imager/index.html

Eight 512x1024 detector panels to create a 2048 x 2048 imaging array. I believe Eneuro had a thought about similar arrays with SEEK microbolometers. High resolution imaging at an affordable price ? Image stitching and optical distortion could be issues though.

Update.

OT.... having just looked at the Bioptic imaging array schematic I suspect my unit has a noisy op-amp or supply rail to such. Each 512x1024 diode array feeds a differential ANALOGUE output to an AD462 amplifier, that in turn feeds the signal to an AD9240. Something before the ADC is injecting noise onto the analogue signal path for that specific array panel. A job for another day though as way too much going on at the moment (plus the Faxitron weighs 70kg and I don't fancy lifting it to get to the base mounted camera!)

Aurora

[blackboxdisease]

is the seek software using apache to show the images, or is the camera using apache to deliver the content?

I ask, because there is a lot of code in the apache directory that looks like it's talking to something.

I also noticed that the camera is struggling more to keep up with the image processing. Could this be due to the extra algorithms? FPS seems lower than the last software.

[eneuro]

There is a single rectangular shutter aperture in one side of the lens 'tube' and the X-Rays are punching through the other side wall producing the contrast where the aperture is located.

Yep, but on the right there is small much more transparent area-looks like no metalic walls on the other side?
Shutter by default isn't inside lens holder, I guess ?

Unfortunatelly, I can't destroy customers Seek dongles, so lets see Seek teardown again.

BTW: Only thermal iron LUT applied on normalized x-ray images-this attached there was additionally blured 2 times 

[frenky]

Progress...

I was able to get rid of horizontal banding on all temperature ranges by subtracting weighted values of 207 column from sensor values.
Higher the temperature lower the weight -> lower the subtracted value.

For me this is good enough. Next step is to test this algorithm on various camera captures. (Must make an USB cable).
After that the real fun begins.
(I'll be mounting module on stepper motor with 1.8 degrees step and then combine multiple images...)

[eneuro]

I was able to get rid of horizontal banding on all temperature ranges by subtracting weighted values of 207 column from sensor values.
Higher the temperature lower the weight -> lower the subtracted value.

Do you mean this this Seek raw sensor data 207th column "rainbow" ? 

How do you know-maybe this is not such simple linear function and... it will change when temperature changes inside this dongle, so isn't better let computer compute those polynominal coeficients in realtime? 
There is nice math behind this approach: wiki: Taylor series

[frenky]

Yes I'm using those values to remove banding. But I also incorporated dynamic scaling so I don't get bright lines on higher temperatures.

[frenky]

The App is starting to look like a pilots cabin.

[eneuro]

Anyway sooner or later this app have to look like Flir's improved thermal interface with F16 cockpit buttons and nice military deep geen

For the moment command line i$ fine 

[DonMcJr]

Hello! I do not know how to do all the cool stuff you guys do...

Just received my Seek  For Android Saturday and was wondering if anyone has the old version of the app? I cannot believe they goofed that bad on the new release and cannot believe they havent fixed it yet...

Thanks!

[miguelvp]

I don't have the previous (1.4.something) , but I have the one before that (Seek_Thermal-v1.3.0.1.apk) here:
http://www.miguelvp.com/ForMike/


Edit: Wait, the previous one is still available here:

http://apk-dl.com/root/apk/2014/11/3/com.tyriansystems.SeekThermal_1.4.0.2_%5Bwww.apk-dl.com%5D.apk

DEHiCKA posted that link in reply #920

Edit again: Argh, cant get the square brackets to escape, still trying.
Cut and paste this while I try to figure this escape sequence out:

Edit, fixed link with escape URL sequences!!! finally!! 
But cut and past the following should work as well

Code: [Select]

http://apk-dl.com/root/apk/2014/11/3/com.tyriansystems.SeekThermal_1.4.0.2_[www.apk-dl.com].apk


[DonMcJr]

Thanks! I just tried it out and the picture looks better on the old version too why did they bother to change it?

[-jeffB]

Just received my Seek  For Android Saturday and was wondering if anyone has the old version of the app?

If you don't mind answering, when did you order, and how long did it take to arrive?

I ordered mine directly from Seek on November 18, and got email on the 19th saying it had been shipped. But FedEx tracking STILL says "Label created", meaning Seek hasn't turned over the package to them yet. I've sent three messages to orders@thermal.com, and I haven't yet seen a single reply (other than "got your email" auto-responses).

I guess I'm just looking for reassurance that they actually ARE shipping currently. From all I can tell, it looks like they're taking money, but not delivering products, and not answering email. I don't want to dispute the charge via my CC, but it seems like that's the next step.