Hi everyone how has same knowledge about thermal cameras

I am geo-scientist and is interested in doing Thermal Mapping with my Skywalker 1880 platform, in order to create thermal orthophotos maps of specific areas of interest. I am planning to make use of a FLIR Tau2 Thermal camera installed in the vertically down position to capture/map the thermal variation of the ground surface or a water body.

I understand that most of the thermal cameras only gives a live streaming video down to your ground control station, from which you can then manual screen capture still thermal images via FLIR systems own software from the video. But in order to create an good orthophoto map one will have to capture hundreds of these still thermal images manually, which would result in hours & hours  spend in front of the computer to manual capture these images, expecially for a large survey area (500ha to 1000ha).

Do any one know of a way: software or hardware configuration which could be able to capture still thermal images at specific time intervals ( every 2 or 4 sec), same as you would do with normal aerial mapping surveys?

Views: 2581

Reply to This

Replies to This Discussion

I'm not sure if this is exactly what you are looking for, but publicLab.org has done a lot of DIY infrared camera work over at- http://publiclab.org/wiki/near-infrared-camera

They have a free online tool for post-processing filtered imagery at http://infragram.org/.

 Most of their work is done by modifying Canon Powershots with a special blue filter and attaching them to Balloons or Kites (http://publiclab.org/wiki/balloon-mapping).  There does seem to be an assumption within the community that using a UAV for such activity would be too technical and expensive but I've been on a mission to prove otherwise.

Hi Andrew sorry for the late reply, but yes i have used public lab’s Infrablue filter on my Canon SX240 camera to capture near-infrared photos. 

It works really well and all the NIR photo on my DIY webpage is taken by this camera. I am now looking into getting an more advance NIR glass filter from Event38 ..or..converting a GoPro to NIR. I will then be using both NIR and colour Gopro’s on my Iris to capture both simultaneously during a survey.

But to come back to the Thermal Camera,  this is not the same as NIR (also called Multispectral), as Thermal is uses to capture heat emissions radiating off a object. NIR senses captures shortwave infrared spectrum (SWIR), while thermal capture long wave infrared spectrum (LWIR).

I have managed to contact FLIR and they pointed me in the wright direction on how to setup their FLIR Tau 2 thermal camera to use it for thermal mapping.

Jaco

Hello Jaco.
As you have mentioned the TAU cameras are equivalent to using CCTV cameras and do not enable you to take still pictures. That being said what you will have to do is devise a way to still capture the live view. if you were to employ a device like the sage tech image-capturing boards which are roughly $1,000 then you could still capture live images coming through by a streaming video. Another alternative method would be to locate a DVR device similar to the horizon HD dvr/camera combo that comes with a camera and a DVR that can be triggered to take still pictures or HD. I suppose if you could use the pen out for video out on the tau you can tap it into a DVR similar to this you would be able to still capture images as needed on the fly. I myself am battling to do this as well and have considered tracing the pen out of the fire wire on the HD video DVR recorder so that I can pump in any video source and do still imagery without having to rely on the horizon camera. if that was able to be done you could use the still imagery that is attached to the aircraft and mosaic it like you would an RGB camera image based on GPS coordinates I believe. by capturing the image and having date and time match the date and time of the GPS data fed from the autopilot you should be able to somehow create a mosaic even with this hazy looking flir image it's hard to say. I figured if I could trigger the camera such via intervalometer it would work. another option that I thought of is using the flir appon the higher end to gamble so that I could hack the app and make it take still pictures via Bluetooth or Wi-Fi connectivity. it would be nicer however to be able to do this through an Adreno circuit or on board and store the data.

Dave Skala
civilianuav.com

HI Jaco, Have you made any progress on this Initiative? I am trying to do the same thing with a TAU 2 336,  Any help is greatly appreciated. 

Thanks, 

Dave

Dave,

Since the TAU can output digital video, it should be pretty easy to pull frames if you can run code on your interface device? If you are taking analog video out, then you will need to run through an ADC of some kind to convert to digital. We are doing this now from Quark cameras, but we use Matlab code on the back end to pull frames. If you are running your ground station on a computer, and if you have a way to convert the analog to digital, then a similar approach might work for you. Another alternative might be to look for Arduino/other MPU code that does the same job. Just curious, how will the mapping software correlate the imagery with the platform state? Based on timestamp of image and log of location data?

cheers,



David said:

HI Jaco, Have you made any progress on this Initiative? I am trying to do the same thing with a TAU 2 336,  Any help is greatly appreciated. 

Thanks, 

Dave

Bcatv 9,

Thanks for taking the time to reply. I appreciate your suggestions. Currently, I am recording live video (I am not sure if it is digital or audio) with an on board DVR, which work fine. However, taking snaps of the video seems to have two problems: 1) The video has an overlay that includes temperature and spot position. And 2) the file has no EXIF data.

I looked at the FLIR setting using the Software GUI but didn't see anyway to turn off the temperature overlay. 

I am trying to create a geo rectified orthomosiac heat map with Pix4d. I know pix4d accepts video but the output won't be geo rectified, which is something I need. And the overlay data would create several problems.

Dave 

What you need is radiometric data.  This at least shows that it is available:

http://www.flir.com/cvs/cores/knowledgebase/index.cfm?CFTREEITEMKEY...

It sounds like you need to look into the Avanced Radiometry function of your camera.
http://www.flir.com/cvs/cores/data/?id=62350

An issue you will likely run into is the automatic scaling of each image based on the temperature of objects in the image. As you move over objects with most IR cameras, everything including the background values change making stitching difficult. 

Corey, 

Thanks. I noticed that in the video I recently took. I've concluded that I won't be able to do what I want to do with a TAU. 

Looking at cooled sensors now like the Neutrino. 

Dave

Reply to Discussion

RSS

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service