DIY Thermal Camera That’s Better And Cheaper Than a FLIR

A few years ago, FLIR unleashed a new line of handheld thermal imagers upon the world. In a manufacturing triumph, the cheapest of these thermal imaging cameras contained the same circuitry as the one that cost six times as much. Much hacking ensued. Once FLIR figured out the people who would be most likely to own a thermal imaging camera can figure out how to upload firmware, the party was over. That doesn’t mean we’re stuck with crippled thermal imaging cameras, though: we can build our own, with better specs than what the big boys are selling.

[Max] has been working on his DIY thermal imager for a while now. We first saw it about a year ago, and the results were impressive. This thermal camera is built around the FLIR Lepton sensor, providing thermal images with a resolution of 60 by 80 pixels. These thermal images were combined with a VGA resolution camera to produce the very cool enhanced imagery the commercial unit will get you. There’s also a 1/4-20 threaded insert on the bottom of [Max]’s version, making it far more useful in any experimental setup.

Now [Max] has unleashed his DIY Thermocam on the world of Open Hardware, and anyone can build their own for about €400 (about $425). The components required for this build include a FLIR Lepton sensor easily sourced from the Digikey or GroupGets, an Arducam Mini, a Teensy 3.6, and a mishmash of components that are probably kicking around your parts drawers.

If you want an overview of this project before digging in, [Max] has a project overview (PDF warning) going over the build. This is one of the better DIY projects we’ve seen recently, and the documentation is fantastic. If you’re thinking about buying one of those fancy thermal imaging cameras, here you go — this one is just as good and half off.

61 thoughts on “DIY Thermal Camera That’s Better And Cheaper Than a FLIR

  1. Sorry, not to disrespect the project itself but the headline “BETTER AND CHEAPER THAN A FLIR” is just wrong.
    In his DIY version he uses the same exact sensor that is used in the FLIR One which is way cheaper. Even if you still need a android smartphone to attach it to you’ll get a solution for under 400€

    1. Although, reading his paper, he claims better specifications such as wider temperature range and greater temperature sensitivity, so the claim is justified (even though, yes, it’s still the same sensor, maybe FLIR are restricting the One for some reason – possibly business/market, possibly stability/accuracy).

          1. im not 100% sure how these thermal cameras work but i know NTCs when pushed too high for too long they can become less sensitive in the higher end
            that might be a part of it
            but more realistically its to prevent the FLIR One from being a direct competition of their higher models ..

      1. The stock FLIR One app limits the resolution and throws in some fake noise. They’ve released an Android SDK though and you can get third party apps instead that don’t have these limitations so you get the full 160×120 resolution. There is a Linux driver as well somewhere on the eevblog forum.

      2. Most of the One’s limitations are in the app on the host device.

        They disappear once you start processing the raw data with third-party Android apps or just a PC with libusb (I don’t have a link at the moment, but there’s a thread on eevblog where people have managed to get access to all of the raw data streams.)

    2. TLDR after I scrolled to a point where they states:
      “The Lepton3 is not available
      on the market as a standalone sensor so far, so it has to be unassembled from purchased FLIR
      One Smartphone thermal imaging camera attachments. The removal is easy and can be done in
      under five minutes, the sensor has a shutter already attached.”

          1. These days it’s getting to be that you can’t even find an article where the editor isn’t shitposting all over the comments. The place is really going downhill.

    3. No, he’s not using the exact same sensor as the FLIR One.

      He’s using the exact same sensor as the *first generation* FLIR One

      The currently available unit is the third-gen – although it appears to have the same sensor as the second-gen, which has is a 160×120 sensor as opposed to this unit’s 80×60 sensor.

      1. OK – can’t edit comments.

        He’s pulling sensors from FLIR One units.

        By definition, can’t be “better and cheaper” than the FLIR One it was pulled from. (The FLIR One hardware does not limit the thermal sensor performance in any way I or anyone else have been able to identify – the app itself is where most of the unit’s limitations lie. But third-party apps like Thermal Camera+ only add a few dollars to the price and still come out WAY cheaper than this.)

    4. The full title is “DIY THERMAL CAMERA THAT’S BETTER AND CHEAPER THAN A FLIR”. If it was only “BETTER AND CHEAPER THAN A FLIR”, as you stated, then you would be correct.

    1. Clearly I wasn’t paying attention when I clicked on this, I started to read it it and thought the same as you, then realised oh it is a Brian Benchoff article no wonder it is just a load of hyperbole. I have managed to avoid any of Brian’s articles or a couple of months. Shame there isn’t an option to exclude posts by a particular author.

        1. “I normally scroll right past whatever I don’t like. But sometimes, I click on something I know I won’t like, because I’m not paying attention. I wish there was a setting so I wouldn’t accidentally make that mistake in the future.”

          “There is. You have the option everyday to scroll past whatever you don’t like.”

          Reedin is hared.

    2. Well the difference is that he is referring to a complete camera and you are referring to the sensor made by the same company, which can be programmed and used in different ways. Hope this clears the misunderstanding for you.

    3. Exactly. As long as it’s still a hundreds-of-euros device (never mind the actual same sensor), please don’t bother waking me up – calling this “cheap” is like the eleventy billion 500-800 euro “cheap” 3D printers that make a big deal of comparing themselves to 1000-2000 euro overpriced tat conveniently pretending the other few billion 500-ish printers never existed.

      1. I fully agree, cheaper got my attention, but then I quickly noticed that the word “cheaper” had nothing to do with cheap as in “it costs almost no money”. $#$#@% they fooled me again.

  2. This is a very nice project from the standpoint of being a true open hardware FOSS project which I applaud. Especially considering it is was homework assignment! Complete schematics, design files for the board, full firmware, etc.

    But as far as making a product which is ‘better’ than FLIRs offerings, It seems like the solution to not liking a companies practices (FLIR) is not to buy more of their products–especially considering there is more profit in selling the bare lepton module than the fully packaged TG130.

    Consider for example the first gen seek compact ($250) has an order of magnitude more pixels (206×156 vs 80×60) and can be directly be used with a computer (using python/libusb, so it will run on anything with a usb port) or cheap smartphone (android/ios). For sub $300 you could easily have a device which is much more powerful than anything based on the lepton, with much less work. For the cost of this project ($400 + assembly time) you could almost afford the gen 2 seek devices (320×240 pixels, $500) and really have a really solid imager.

  3. Hey guys,

    I am Max, the creator of the DIY-Thermocam.

    The title of this article may be sensational and I agree that other solutions for a comparable (Seek Reveal XR) or lower (FLIR One) price exists. I am just a single student who made this project in his free time, so I can not compete with big companies like FLIR or Seek with hundreds of engineers and a big budget. I think the title comes from the comparison table on my page, where I contrast the specs to a FLIR E6, costing around 2000€.

    The added value of this project is certainly the software part. The open-source firmware has a lot of nice features and you can control it over a touch interface. The memory is replaceable and the device can be connected to a PC or Pi for live thermal streaming. There is also a nice thermal editing software called “ThermoVision” from JoeC, that allows you to edit the raw images like the FLIR Tools do.

    I put a lot of time and effort in this project and hope you do not only judge about it by reading the title. It may not be the cheapest solution in the meantime, but for that price, you can make your own nice thermal imager, and you can adapt it to your own needs at any time.

      1. Yes, I totally understand that.
        I did not choose the title, but I am happy the project is at Hackaday at all, so interested people can find it. Then they can decide whether they prefer the FLIR/Seek solutions or the DIY-Thermocam by their own, which is always a tradeoff decision.

    1. No worries. I saw many readers ranting about the sensational headlines of this website, But it probably has a direct financial impact on the website through traffic increase and ad revenue, so probably the administrators of this website keep their head buried in the mud…

    2. As you point out, it isn’t competitive on price since you need to buy a lot of extra modules but it is a very well implemented/documented project and good experience. If you’re interested in going into the electronics field for work then this would look great as an example of what you can do.

      I did previously think about connecting a Lepton module to a RPi Zero + RPi camera and a SPI LCD but the price of the Lepton + breakout board was always too high. There are two SPI interfaces accessible on the expansion header and the I2C interface is unused as well so you’d be able to drive the LCD and Lepton together. Although if you really want to go the easy route then you might as well just plug in the FLIR One for Android module and skip the separate camera module as there is a reverse engineered Linux driver somewhere on eevblog. Most of the limitations like reduced thermal resolution and fake noise are added by the FLIR One app itself.

    3. oh the project itself is completely OK, it is the title of the article that is the problem. Rest assure that this happens all the time and nobody will hold the creator of the project responsible for it (but the hackaday author). Please keep doing projects like this.

  4. 80×60 just seems not of this time, and it’s not a technical limitation but an artificial one.

    No offense to the hacker who made this though, always good to tinker, but rather it is a critique on the US companies that make such modules.
    And even the author says that to get a lepton3 module you should but the device for smartphones and rip it out from there, which shows you how artificial it all is.(not that that 160×120 is so thrilling either.)

      1. You can get 4 times the resolution from Seek for around 400
        And seek supports a wider range out of the box.

        The prices mentioned in the PDF of the author are really odd incidentally, they are like references from 20 years ago that somehow slipped into 2017. So don’t blame HaD for that people.

          1. The c2 is based on the lepton and is only 80×60. The LCD screen on it is 320×240 but that is only for the visible camera. The seek pro is actually 320×240 thermal pixels. Cheapest products from flir that has actually qvga resolution are a few thousand dollars.

            In general seek has been outperforming flir ever since they released the first gen seek just before the lepton was released. They started out as some disgruntled flir employees fed up with the crap that flir tries to pull.

    1. this looks like a fantastic deal … until you actually see pictures it produces :( for more info come to eevblog
      Seek might be good with your own filtering/smoothing of raw data from the sensor, but as it is looks like chinese USB 1.1 200K pixels webcam interpolated to 1.2Mpix.

  5. Just to make sure everyone is clear, this article talks about using FLIR sensors from FLIR One cameras. This means that by default it is more expensive than buying a FLIR unit. Also, the current Gen II model of the FLIR One provides a 160×120 resolution (19,200 detection points) and costs $250. That is 70% cheaper than this “upgrade.” If you want more functionality with the files there are plenty of 3rd party software makers in the Apple and Android app stores, or you can create your own as the hardware is not locked down to restrict independent software development.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s