Image Quality Comparison of FLIR Tau 2, Boson, and E60 LWIR Thermal Cameras
Вставка
- Опубліковано 8 жов 2024
- A slightly rambling comparison and discussion of three thermal cameras by FLIR:
A Tau 2 640 (30Hz, non-radiometric, commercial sensitivity grading, analog video output VPC, 32deg HFoV)
A Boson 640 (60Hz, professional sensitivity grading, UVC output VPC, 50deg HFoV)
An E60 (60Hz, radiometric, standalone with USB streaming output and Composite output pictured, 25deg HFoV)
Includes discussions of adjusting some settings, notable quirks of the cores and software, and some demonstration of palettes and other features.
Such unique video!.... Where did you get them, I even didn't find those models on sale. What do you think about Vue models (found on Adorama), probalby old and in 2021-ish are there better options?
I have generally gotten mine on the used market, and at least inside of the US there is some availability on ebay and the like (>9Hz framerate cameras are export restricted from the US and there are similar laws elsewhere). From a quick look, the Vue probably uses an ISC0901 sensor, which means the image quality is likely comparable with the FLIR Ex series cameras (not the Exx series pictured here), but since it's fairly popular in UAVs, you can probably find some video from it around.
@@mdbssn Hi, do you know about regarding registration, like you can't export abroad, but do you need somewhere register 30-60hz devices? Or it's no needed registation for use inside the US (as a private person, for nature observation etc. simple civil activity)? And, will I able to sell it later on ebay inside the U.S.? Is there regulations?
How did you get the video output to the PC to record this? I have a Tau 2, but the software I use to hook my Tau 2 (with a VPC) seems to only allow me to take snapshots and make setting changes. The mini coax port on the VPC died after using it for roughly an hour, but the VPC still powers on and connects to the computer. If I could use the USB connection to the computer for live feed instead that would be great.
You have to pay a pretty hefty premium for the digital output, I think it's only the CameraLink VPC from FLIR, but there are USB ones from third parties. That said, I used a cheap USB analog video capture dongle to capture this. Most options should work, but my preference is one that uses the UVC standard because it integrates well with everything without drivers most of the time.
I remember reading that in atronomy, they'll take data from multiple cameras and digitally combine them to get better resolution, is that something that could be done with these cameras to see better detail from further away?
Definitely, and I know a few FLIR and Keysight/NEC Avio cameras that include a mode for this, even using a single camera sensor. The disadvantage is that it both takes multiple exposures and some amount of post processing - so unless you have a very fast readout sensor, you're not going to get realtime video out of it.
There could even be something similar to this happening in the Boson already - it uses a good bit of image processing hardware onboard the camera, and I think this is why the sharpness adjustment in particular can achieve some very high resolution appearing images.
You can stitch images together to create a functionally higher resolution image but what they do in astronomy is some science magic involving interferometry that makes it so their image quality is effectively what it would be if they had a lens the size of the area between them, not just the sum (at best() of the composite image. We're only now doing this with visible light in astronomy because their wavelengths are in the hundreds of nanometers, which is the precision you need in placing your receiver/mirrors from two telescopes to get that benefit. We've been doing it with radiowaves for a long time though since their wavelengths are huge. That black hole image was taken using digital images timestamped by atomic clocks at radio telescopes around the earth, which allowed us to create what was effective an earth sized receiving dish. Though, to clarify, this only increases image quality, not brightness.
So which camera is the best per say for a vehicle to see at night at distance? For things like animals people, or per say if someone was pulled over ahead of you with his lights off and radar off? You know like seeing the heat sig of a running motor in the pitch black.. which would be the go to camera? I guess they're discontinuing the current "vehicle" set up. I'm just curious if there's a decent flir camera for being able to see things at night in real time.. any suggestions?
I doubt any of the cameras used here alone would be the right choice. If you're looking for detail at distance, you want a narrower field of view, and while there are camera cores with longer focal length lenses, they aren't integrated with a screen, aren't waterproof, aren't vibration resistant, etc. For simple spotting, maybe a spotter scope is a good choice, but otherwise getting a camera model that has been weatherproofed for outdoor use is probably a good bet - I haven't worked with any personally, but they do build these cores, and cores like them, into marine camera systems (often turreted, like a security camera), then there are security cameras with a thermal camera installed, and there are the aforementioned hunting/spotting scopes.
It's not that these couldn't work, it's just that there are likely better lens/packaging options for something like what you are describing.
@@mdbssn that's my conundrum I'm under the understanding that thermal doesn't work very well through glass? I suppose that makes sense as it'd be a heat soak but something I didn't consider. I've also been looking at drone cameras as they'd have shock/ weather resisting and some have thermal/ zoom capabilities.
One of your tested cameras was listed as the "next in line" for flir to use as these thermal vehicle cameras but its hard to find any actual testing. Some are claimed to have heat detection to 1500 yards while others are significantly closer at 550 ish yards I feel like 550 is more realistic then the 1500 claimed
Is the best way to acquire a development kit and thermal core to go directly to the FLIR website?
I don't know about development kits, but cores can be bought direct from FLIR, from authorized dealers, or on the used market (but high framerate cores are international export restricted from the US, so there are potentially some rules to deal with). The VPC (video, power, control) module to read data out of the core can be done similarly, but I've typically gotten mine from FLIR directly because they aren't as available elsewhere.
I don't understand why anyone would test a thermal to see a target 5 meters away. The naked eye could see that as well. We need to see if the thermal can detect 500 meters etc. Testing it in conditions where the human eye can't see. It's like testing night vision at noon.
Night or long distance vision isn't the only nor primary application of long wave infrared cameras. These devices visualize a band of the EM spectrum that the human eye simply can't, so the ability to visualize it at any distance augments the perceptive capabilities across the board, and there are many applications where thermal analysis of a specific region - far, up close, or even at microscopic or telescopic scales - is important to understand heating without contact sensors.
While I don't make the most interesting subject for such cameras, the fixed focal length lenses on most thermal cameras mean they're well suited for a specific range, an in the cases of the ones demonstrated, it's this near to mid range (more like 3m away in this demonstration, the e60 is the best suited to longer distances but still has something like a 25 degree FoV.)