Flir E4 Thermal imaging camera teardown

[stefbeer]

Just for fun (and maybe because I'm a little stats-freak ) I made a graph for the posts per day in this thread. See attached image.
I used the dates according to my timezone setting (UTC +1), so if anyone wants to try this at home, he / she may get a litte bit different numbers.
Oh, and just for the record: October 24th was the day when Mike confirmed the 320x240 Pixels resolution

[Taucher]

Can someone please tell me what progress has been made since the 3b resolution hack and taucher's beta2 mod? I saw reference to a zoom function etc. Are these features going to be added to the menu hack?

I'll most likely un-comment zoom in Beta3 - but it's just digital zoom, there's no image quality to gain - so not urgent
Right now I'm pretty busy and I'm lacking enough free time to play with the E4 

[Fraser]

Taucher,

your efforts are greatly appreciated and I know exactly what you mean regarding finding spare time to 'play'. I have been using my E4 a lot recently and I must say I welcome the Zoom capability. The 45 Degree lens FOV is normally specified for close-in viewing such as a small room with an electrical distribution cabinet. The common 'standard' FOV on industrial thermal cameras was 24 Degrees. I am used to such, so find the E4 a little too wide angle for some tasks. Targets appear small on the display and so can be harder to interpret.

You are quite correct that the Zoom feature adds no improvement in resolution, in fact it reduces such. What the electronic zoom does offer though, is a larger target image on the E4's built in screen in real time and without the need of a PC. This can be useful where the target is small in size, such as a PCB SMT component. The ZnSe Close-up lens gives an excellent pcb inspection capability, but it will be nice to be able to zoom in closer when required.

The X2 zoom is likely to be my most used zoom level as 160x120 resolution is acceptable for my needs, and provides the approx. 24 Degrees FOV that I miss in the unenhanced E4. 

[Navynuke]

Happy New Year! Today I got a Google alert that the K-Series (fire fighting version of the E40)  trade-in program was extended. Not something I was particularly interested in except for the fact that the K series unit has two temperature ranges -20C - 150C and 0C - 650C as well as a fire palette that looked pretty cool. To that end  I wanted to see if there is anyone who has stumbled across any of the other temperature ranges in their exploits in other Flir cameras (Exx series has multiple ranges and I remembered someone renting an E60 for a short period of time) I was thinking that if we had the files from one that the higher temperature range might be made available and calibrated through the astra web interface. As it is right now only the one range is available. I have some applications where a higher temperature range would allow me to NOT apply too much heat to smaller fragile items and is on my "Thermal imaging wish list". Thanks for any thoughts consideration...

The K series info is available at: http://www.flir.com/uploadedFiles/CVS_Americas/Law_Enforcement/Products/K-Series/FLIR-K-Series-brochure.pdf

[daves]

Which procedure do you have selected?
The image is a stitched msx panorama and not a Flir radiometric image (set measured points in Flir Tools)

Since FLIR makes JPG compressed too much, it makes ugly smudges and you do not know what is heat and what is jpg artefact :-D. So I use this procedure:

Extract IR images from all parts of panorama:

Code: [Select]

>panorama.cmd
I am using modified version of panorama.cmd, to produce visible IR:

Code: [Select]

@echo off
for /f "usebackq delims=" %%f in (`dir /a:-d /b /oN FLIR*.jpg`) do call :over %%f
goto :eof
:over
set a=%1
echo %a:~4,4%
exiftool -b -RawThermalImage %1 | convert - gray:- | convert -depth 16 -endian msb -size 320x240 -level 10000,16000 gray:- RAW%a:~4,4%.tif

Then I extract also all real images:

Code: [Select]

>exiftool -b -EmbeddedImage FLIR*.* -w Real_%f.jpg
Now is time to stitch IR and REAL images, so I will have two panorama images.

After that, I convert panorama back to nonvisible IR:

Code: [Select]

>convert Stitched_panorama.tiff +level 10000,16000 stitch.tiff
And finally convert to some palette:

Code: [Select]

>convert stitch.tiff -level 10520,10900 pal.png -clut test.png
Then I take some image editor and put IR over REAL. Result you have seen before.

[pomonabill221]

Can someone please tell me what progress has been made since the 3b resolution hack and taucher's beta2 mod? I saw reference to a zoom function etc. Are these features going to be added to the menu hack?

I'll most likely un-comment zoom in Beta3 - but it's just digital zoom, there's no image quality to gain - so not urgent
Right now I'm pretty busy and I'm lacking enough free time to play with the E4 

HAPPY NEW YEAR!!!!

Yes just digital zoom but it appears to have some image smoothing so the quality isn't that bad.  I also removed the X8 zoom as it was really not very usable, even X4 is pushing it but could be useful in some situations.

[plesa]

Happy New Year! Today I got a Google alert that the K-Series (fire fighting version of the E40)  trade-in program was extended. Not something I was particularly interested in except for the fact that the K series unit has two temperature ranges -20C - 150C and 0C - 650C as well as a fire palette that looked pretty cool. To that end  I wanted to see if there is anyone who has stumbled across any of the other temperature ranges in their exploits in other Flir cameras (Exx series has multiple ranges and I remembered someone renting an E60 for a short period of time) I was thinking that if we had the files from one that the higher temperature range might be made available and calibrated through the astra web interface. As it is right now only the one range is available. I have some applications where a higher temperature range would allow me to NOT apply too much heat to smaller fragile items and is on my "Thermal imaging wish list". Thanks for any thoughts consideration...

The K series info is available at: http://www.flir.com/uploadedFiles/CVS_Americas/Law_Enforcement/Products/K-Series/FLIR-K-Series-brochure.pdf

The firefighting mode skin ( green numbers and smaller temperature scale) looks much more better than original one. I will really appreciate to have it on Ex camera enabled.
I'm also interested what is the difference between various temperature range cameras. Is it only calibration?
It seems that the zoom feature is for this version Kx0 and not for Ex series.

[Fraser]

Just installed activated the Electronic Zoom menu....... Excellent 
A very useful facility to have on board.

[tomas123]

Which procedure do you have selected?
The image is a stitched msx panorama and not a Flir radiometric image (set measured points in Flir Tools)

...
Then I take some image editor and put IR over REAL.

interesting, a fast way and your images overlay looks great

I tested some time ago Hugin. But Hugin is hard to control.

Hugin Panorama
...
With minor changes in the flir.php script you can save real and thermal image as two images with the same pixel size  and the same field of view.
Now you can stitch the real images with Hugin and save the *.pto project file.
In the next step load the thermal images and use the *.pto with the real image control points for stitching.
After then you have a real and a thermal panorama with the same field of view (make a MSX with IM convert etc).

This method with hugin is great for thermal images with minor difference in temperature ...

... where you can't find good control points in thermal images

[Bonocr]

Hi to all,
my first post here.

Just hack my new E4 (Mike hack and ADDMENU_BETA2a), it works wonderfully!
Thanks to Mike and other great people here!

For log purpose:

Cal date: 17/Dec/2013
SN:63910xxx
Model: E4 1.1
SW: 1.19.8

[Fraser]

Thanks for s/n & FW info.

While the process is still fresh in your mind, it would be a good idea to add the Zoom menu as well. Details recently in this thread.

The complete FlashBFS file is available to use but I decided to edit my current Menu2b file instead.

I had some issues editing the XML file with the MS 2007 XML notepad due to 'illegal' characters in it, but ordinary notepad worked fine on the XML and allowed enablement of the zoom functionality.

[Bovvy]

Hello all, new boy on the forum. I discovered this thread just before Christmas and have finally got through all 179 pages, God knows how because after page 100 most of it went over my head!

I have held off getting the E4 so far, but now thanks to the information in this thread I am pretty sure that I can apply Mike's upgrade and unleash the E4's full potential.

I have a trip to New York approaching and wondered if anybody could recommend a reputable dealer there, or one which would deliver to an address over there. Obviously I will Google this, but a recommendation would gratefully received.

Secondly, did I read that the E30/E40 is also upgradable with E60 firmware?

Thanks in advance

Bovvy

[Bonocr]

Thanks for s/n & FW info.

While the process is still fresh in your mind, it would be a good idea to add the Zoom menu as well. Details recently in this thread.

The complete FlashBFS file is available to use but I decided to edit my current Menu2b file instead.

I had some issues editing the XML file with the MS 2007 XML notepad due to 'illegal' characters in it, but ordinary notepad worked fine on the XML and allowed enablement of the zoom functionality.

Thanks Aurora for your answer.
If I am not wrong for the zoom features I have to uncomment some rows in the xml file and reload it by FTP, is it right?

I have also printed the tool for focusing and the lens holder see the pics.

[Fraser]

Hiya,

Nice 3D prints 

I have not seen that focus tool before. Can you share the STL file ?

You are correct regarding uncommenting the zoom statements in the XML file. You will also need to set the zoom statement from false to true in Mikes resolution upgrade file. This will require that you also run the CRC calculation again. Its the same process that you carried out for increasing the resolution but you need to first make the change to the zoom configuration line. There is no need to remove either of the files that you have already installed. Just over write them via FTP 


I have attached the XML file, within the FlashBFS folder structure, that I used. It is using the Version2b baseline created by Uup (which uses the XML file originally created by Taucher)

I have also added the edited resolution file with the Zoom option set to true

Update. I think I attached the wrong XML file so I have deleted it and added the right one. Sorry.

Update. I have added the Uup menu file set called FlashBFS that I edited to activate the Zoom function. Use this if you find it easier than using the XML file alone.

[QuantumLogic]

Hello everyone. I just received my E4 from Valuetesters.

Cal date: Nov 21, 2013
SN:63907xxx
Model: E4 1.0
SW: 1.19.8

(Info matches another already in Aurora's list)

[daves]

After I successfully added the zoom option, I was messing with calibration file calib.rsc, since cold values are off anyway, so it could not be worse.

I used values from excel published here without any modification (because it cant be worse anyway) and I have great results.

Freezer is no more showing -40°C, but -18°C now. Ambient temperature is shifted +2.5°C - I believe when I do correct tests to get my own calibration values, I will get yet better.

[john19]

After I successfully added the zoom option, I was messing with calibration file calib.rsc, since cold values are off anyway, so it could not be worse.

I used values from excel published here without any modification (because it cant be worse anyway) and I have great results.

Freezer is no more showing -40°C, but -18°C now. Ambient temperature is shifted +2.5°C - I believe when I do correct tests to get my own calibration values, I will get yet better.

I accidentally got the wrong sign in the error columns. It doesn't matter much, but it may look confusing. Column C should be "B2-A2" etc, not "A2-B2", and column F should be "E2-A2" etc. It will still produce the right calibration parameters, but I'll attach the corrected sheet here anyway.

[daves]

I accidentally got the wrong sign in the error columns.

I have noticed, but I though you meant it that way.

By the way - I had no time to deep tests yet, but seems to me if you put camera into cold, it shifts whole range down. I mean - I keep camera at 20°C and point at matte black painted steel and it says 23°C. I take it out in 0°C for a while to cool down a bit and take it back at 20°C and point again on that black steel and it says 19°C. If I let it warm, it gets back to 23°C.

What I am trying to say is, that the line probably should not be flat (horizontal), but should be tilted to count with temperature shift. Just guessing.

Great job anyway, It was kind of pitty to take camera out while it was telling you its -11°C or so, when actually was just below 0°C.

BTW: I do not know If someone told it before, but J0 = 1/R2

[john19]

BTW: I do not know If someone told it before, but J0 = 1/R2

You mean J1=1/R2.
J0 is the same as O in the excel sheet although in the calib.rsc file the sign for J0 seems to be reversed.

[Fraser]

I have been playing with the XML file.

It is very easy to change the menu options present and the order in which they appear 

I have used the 'Zoom enabled' XML file produced by 'freak_ge' as the basis of my own personal XML preference.

I changed the palette order to the following:

Top row: Iron / Rainbow / Rainbow HC / White hot / Black hot / Arctic / Lava
Bottom row: Blue below / Red above / Interval / Detect Condensation / Detect insulation deficiencies

If anyone else wants to use this version of the XML, I attach it for your use. Just take the FlashBFS folder (after unzipping it) and use FTP to overlay the cameras FlashBFS folder.

[QuantumLogic]

I know this has already been discussed before but I must say that the USB port is the absolute worst design.
Everything else about the E4 is beautiful, but that plug..... 0/5 stars on that.

Big props to Mike (researcher and high-rez enabler), Taucher (menus), freak_ge (zoom) and everyone else I missed for the discovery, subsequent learning and implementation of useful feature additions to the E4 to make it even better.
It is an incredible tool.  I've already found where the air ducts route in the floors, ceiling and walls as they go to the various rooms.  So cool!

I noticed after just a short while of using the E4 that the frequency of calibrations slowed down a lot.  I mean, after turning it on it was calibrating once every minute or so, but after being on for a while it only calibrates once every few minutes.  Definitely not as annoying as I first thought it would be.  Once the internal temperature stabilizes I guess it decides to reduce the frequency of calibrations, which makes sense.

I would not have purchased an E4 if it were not for the enhancements discovered here.

[Fraser]

Yes the Micro USB connector has been highlighted as a very poor choice for what is otherwise a rugged design. I personally hate it. My eyesight is not what it was and I am always struggling to see the correct orientation of the plug and then getting the angle right for entry into the socket...... really cr*p 

For info the external charger uses the same micro USB power connector and so can use the plug pack power supply supplied with the camera.

I reported my disappointment to the CEO at FLIR, and suggested that they might wish to revisit that aspect of the design in any future hardware updates.

[Fraser]

Additional FLIR Palettes inc Medical !

I have been looking at my file archives for the FLIR PM series of industrial cameras and have found a pile of palette PAL files. I have no idea whether these can be used with the E4 or will need conversion.

I am attaching them for experimentation only as some may be of interest:

These are common industrial palette types plus some odd ball palettes like 'Medical' that is used to diagnose breast cancer and other thermally detectable illnesses.

I suspect that the headers will be wrong and will need adapting to the newer PAL format ?

For those wondering what glowbow palette looks like.......



Enjoy 

[Taucher]

Additional FLIR Palettes inc Medical !

Any examples how they look?

[Fraser]

I can produce some test images in the different palettes. Some may easily be found using a Google search as they are Industrial standard palettes. I'll see what I can do on example images

Having compared the PM Iron file with the E4 Iron file, the headers are totally different and the data blocks differ significantly as well. These PAL files may be of no use at all but I thought I would release them anyway.

Medical palette is detailed in the attached FLIR document on vet use.

This url provides examples of medical use on humans:

http://spectronir.com/image

This is also an interesting video on medical palettes