How can I use the FLIR Lepton and breakout board with UAV FPV?

1,101 views
Skip to first unread message

Isaac Aaron Piche

unread,
Sep 18, 2014, 1:23:19 AM9/18/14
to flir-...@googlegroups.com
Hello,

Like most FPV pilots, I'm using a gopro with either a 5.8 ghz video system or 1.2/1.3/2.4GHZ etc. and I want to integrate this low cost system into my UAVs for search and rescue operations, among other applications. I looked at the pictures of the board, but I have no clue how to go about integrating it. Can anyone provide me with some guidance please?

Pure Engineering

unread,
Sep 18, 2014, 8:14:13 AM9/18/14
to flir-...@googlegroups.com
There have been several requests for a small stand alone product ready for use rather than a development platform. 
The current breakout board requires the use of a micro-controller to turn the thermal camera into an end product. 
I am in the middle of making a small camera system using the flir lepton, it would be nice to have some requirements for exactly what you need for UAV FPV and aerial photography.  

These have been the requests so far:
Ability to take photos and save to microSD
RCA video output 
Ability to power and trigger photos from RC servo line
Ability to survive crashing
Lightweight 

If there any other requirements/requests, please post them here and I will try to incorporate them into the design.  
Message has been deleted

Isaac Aaron Piche

unread,
Sep 18, 2014, 11:37:02 AM9/18/14
to flir-...@googlegroups.com
This sounds great! The ability to save the footage to a micro SD card is definitely important, as well as an RCA video output. The ability to survive crashing and lightweight I will second as well. If the camera housing included a pan and tilt system and/or stabilizer you would have a serious product in the market at a price point anywhere under $1,000 and a unit that any SAR drone operator would need. Pan and tilt is not necessary though, for me just the ability to wire it into standard video RCA style components is a must.

I'm also interested in developing software to recognize thermal images and distinguish between horses and cows, as well as people and other heat signatures, and integrating it into a smart phone app.

How soon should your system be ready?

Isaac Aaron Piche

unread,
Sep 18, 2014, 11:37:59 AM9/18/14
to flir-...@googlegroups.com
This sounds great! The ability to save the footage to a micro SD card is definitely important, as well as an RCA video output. The ability to survive crashing and lightweight I will second as well. If the camera housing included a pan and tilt system and/or stabilizer you would have a serious product in the market at a price point anywhere under $1,000 and a unit that any SAR drone operator would need. Pan and tilt is not necessary though, for me just the ability to wire it into standard video RCA style components is a must.

newir...@gmail.com

unread,
Oct 17, 2014, 6:35:49 PM10/17/14
to flir-...@googlegroups.com
How soon do you think it will be ready? Put me down for one. Also it would be great to have it in a  case that has a micro CS mount for lens. For me just composite video out that is standard for CCD cams is all that is needed. The xmitter will get the video to the ground for SD recording.

Pure Engineering

unread,
Oct 17, 2014, 8:41:22 PM10/17/14
to flir-...@googlegroups.com
The most expensive part of a thermal camera is typically the lens. Regular glass and other normal lens material does not allow the thermal light to pass through it.  

Also generating color composite video from the digital image requires a fairly fast DAC and micro controller to run it. it would possible to generate 64-256 level B/W composite videos though, how useful is this?

Its actually much easier to save the video locally to a microSD. If you can save the video to a microSD, how useful is composite 64 level B/W thermal video?
 

newir...@gmail.com

unread,
Oct 20, 2014, 1:53:37 AM10/20/14
to flir-...@googlegroups.com
Lens....I will look into that.....see what others have been using. What ever they use it has been around for a long time and should be affordable.....20 plus years?

As far as color...I didnt realize you could get color.. real color anyway from a microbolometer. The color wouldnt matter to me I just need to identify an object in total darkness or just to know there is some heat producing object, not necessarily identify what it is, such as an object in the distance moving or a crack in a door jam that indicates heat loss or cooling loss. I realize the resolution of these are low but it is better than nothing. They can only get better from here.

The idea is to see an object in real time to identify the location at that moment. (Flying in UAV first person view..live feed form quadcopter) So all I and most anyone would need is a composite video feed out. A  1 volt PP composite ( or B/W)  video feed I think is standard and can be used by DVR's, computers, web cams, security cams, hand held TV's  and monitors.......with the RCA jacks coming out of it.
A video feed could also be used to get still photos from. You have a better chance of getting that one great shot / photo if you take many shots. 
..taking photos, security cams, heat loss in buildings, heat problems in electrical circuits, sub stations, electrical panels, it could be recorded with any recording device, there are many, many applications for a video feed, UAV being just one of them. But transmitters for first person view in UAV's are mostly used with composite / B/W  video. So it would be better if the conversion was built into the cam. Also from what I read the frame rate is only 9 hz. That is a really slow . Or am I looking at that wrong?


http://www.sofradir-ec.com/landingpage-sensors_show.asp   --   these also output the 14 bit digital and sell for $1500 each.    If you could make something this size and output a standard composite video signal at a price around $400 or less you wouldnt be able to make enough. You would soon have to out source them. 

Thanks and I will look into those issues.



 I will research that and thank you!

John Kuczera

unread,
Oct 24, 2014, 7:25:12 PM10/24/14
to flir-...@googlegroups.com
I just started looking into this today as someone went missing in my remote community yesterday, and it seems little effort is made at night. A live video feed on a device with autonomous flight would allow for quite a search radius in a short period of flight.

I would join the group buy for the Flir Lepton if it wasn't for the insane customs fees associated with a device like this. It would actually end up cheaper for me to get a Flir One from the apple store here in Canada.

Any idea when the breakout boards will be available again?

John x

unread,
Nov 29, 2014, 8:58:05 AM11/29/14
to flir-...@googlegroups.com
Hi,
I'm new here, this is my first post.

I'd like to be able to save still photos to SD card, say 1 or 2 per second.

Live out via composite video would be good too.

Lightweight, weather resistant and affordable would be great.

Thanks,

John

tz

unread,
Nov 30, 2014, 11:11:52 AM11/30/14
to flir-...@googlegroups.com
The Raspberry Pi can do that, even the model A as far as writing out to the SD card, though you might want to repartition it so that the first (DOS/FAT) partition is as large as possible and has a DCIM directory and a subdirectory, then just convert and store jpegs there.  In one of my sample programs, I'm writing out "raw" grayscale PNGs in one of my programs and I give a command to enhance them so you can actually see something.  You can add another image-magic command to convert the PNG to JPEG.  (note to self: shift the 14 bit data to the left by 2 bits so there might be some contrast even in the raw, at least a fire on snow scene)
Reply all
Reply to author
Forward
0 new messages