SparkFun Forums

[rppearso]

Hello All, I am new and was recommended to this website by faculty at my university. I have a BS in chemical engineering and am currently working on the undergrad pre-reqs for a masters in EE. I would love to build a forward viewing IR scope. From what I understand this site sells DSP/microcontrolers with multipule analouge inputs with built in LCD screens. Do you sell microbolo meters here? Do you sell optics/lenses or know where I could order them or the scope case or where I could have a shell made.

The reason I need 2 inputs is because I want to display an IR image as well as the visible light image (overlayed on each other), there will be some challenges on the sensing device side to figure out how im going to do that without having redundant optics. Since IR is very close to the visible spectrum I am hoping I can use traditional visible light optics to focus IR along with the visible light without any/much distortion in the IR zooming/focusing, then the challenge would be to duplicate the image for 2 different sensing devices. Can the DSP/microcontrolers be battery operated?

I can program in VB but it would not be a stretch for me to learn C or C++ if nessicary, but VB would be ideal. I am also eventually going to take a DSP class and possibly a VLSI class in the future as well as some optics classes.

Thank you for the help, I am looking forward to the parallel journey into a masters degree and doing some cool things with electronics.

[waltr]

CCD cameras are sensitive to the near IR spectrum (check the data sheet for the camera sensor element for the spectral response). These are available in a wide price range from WebCams to Lab grade. Most do have an IR blocking filter that once removed images IR very well. Other filter can be placed in front of the camera to exclude visible light and/or select a range of wavelengths.
A good source of filters, optics and lab grade cameras is Edmund Optical.
http://www.edmundoptics.com/

VB programming is really only used on PCs as it is a MicroSoft language. The development tools for micro-processors and DSP processors either use assembler or C for the most part.

Do you sell microbolo meters here? Do you sell optics/lenses or know where I could order them or the scope case or where I could have a shell made.

This is a user's forum so queries to SparkFun should be made directly.

[Cannibal]

From what I understand this site sells DSP/microcontrolers with multipule analouge inputs with built in LCD screens.

Welcome to Sparkfun.

While you're right that this place does sell a lot of nice stuff, the realm of image manipulation is outside of what most embedded controllers are designed for.

If you look up a robotic project called the Surveyor-1 there is a stereo vision module that has two cameras offset for 3-d vision where each camera has a ~500 MHz Blackfin DSP available here http://www.surveyor.com/stereo/. You might be able to modify this system if you can add/remove visible/IR filters from one of the cameras each, but I think you're still in for learning some C/Assembly. Additionally, you'll have to be careful with what resolutions you want to work with - if you want more than what these cameras can do (1.3 MP) you may have to abandon the embedded aspect and look instead to post processing on a PC.

[rppearso]

I dont mind learning a new programming language but I do want to keep costs under control, typically pre-packaged IR devices run around 10k which is outragoiusly cost prohibitive when you can buy a microcontroler with LCD screen for 70-100$ and a microbolometer for 110$, the scope case and lenses might be a little more but not 10k more or anywhere close. I want to avoid buying something "out of the box" and having to mess with it, I am willing to take the time to build up the device so I get what I want.

I would need VERY high resolution, I will eventually want to make this into a long range scope where I can silouete a heat source image on top of a visible light image at maybe 1000 yards. I have seen visual outputs for various things on the market and the resolution is very poor and I know that much better can be attained, ever see the movie preditor and that was in the 80's.

Are there special forums to talk to the sales people at sparkfun to help me find the right parts I need. Thank you for the links for optics, I think the optics part may be pretty tough, but not impossible.

[waltr]

Go to SparkFun's home page and click on Customer Server. There you will find the info on how to directly contact SparkFun's staff.

[MichaelN]

I'm not sure if you realize the complexity of what you're asking for. It would take a team of multi-disciplined experts lots of money and time to develop something like this, and in the end you'd be better off buying something off-the-shelf.

There's no way you'll be able to adapt optics from a normal camera to do the job. It simply will not pass the mid-wave (3-8um) or long-wave (8-15um) IR that you're looking to use. The movie "Predator" used a long-wave IR camera, whereas the typical vision you see from police helicopters (ie, FLIR) is mid-wave.

If you're really keen on doing this, I suggest you start looking for surplus imagers. I think I saw some surplus mid-wave units a while back that were originally intended for a driver-assist "night vision" safety feature on high-end cars.

As for resolution, there's no way you're going to do high-resolution on a budget. Even 640 X 480 is probably way out of your range.

[Roko]

Out of curiosity, where can you find microbolometers in the $110 range, that meet the "High resolution" requirement?

The best I've seen in that range are a either a single element, or a 1x8 array

[tronic]

I'm a bit confused. Are you looking to build a camera that can see in the dark by illuminating a scene with infra-red light? Or are you looking to build a camera that can see heat signatures emitted from warm bodies? The former is much easier to build than the latter. The optics to build a camera that can see heat signatures has to be kept cool using liquid nitrogen.

[tronic]

Ah - I just looked up Microbolometer and I see they dont need cooling. This is a different technology from what I've seen.

[rppearso]

I have not looked at the details of the varying microbolometers, higher resolution devices may be more expensvie but not 10k expensive, its not that im on a budget per say im just not willing to spend 5 and 6 figures on a imaging device, not because I dont have it (at least 5 figures anyways) but because I think its extortion espeically when the individual components are so cheap (relativly). Plus I am doing a MS in EE in parallel and plan on committing 5-6 years to both the parallel paths of the MS and this project and this scope will probably be over 1k but I am hoping it will have supurb resolution.

Anyways I am looking to build a device that will sense and display warm bodies. As far as optics I plan on installing them from scratch and modeling the set up of a rifle scope, I may have to order the lenses that can focus IR long wave radiation, probably not cheap but not 10k either. I have a link on my home computer for the site to order microbolometers I can post when I get home and then I will just need a techs help to get the right microcontroler/LCD screen that could interface with it, from there I would have to start learning about optics maybe order a text book (that I would need anyways for a later masters class).

I do need to learn the resolution lingo though, like what kind of resolution/array size would you need to intricatly silouete a warm body from 1000 yards, or heck even 600 yards

[04kmorri]

I wish you luck on build your FLIR scope, but I think using a CCD camera and and microcontroller at best may only yeild the standard night vision which suffers from (overexposure?) if you happen to look at a light source, where as true FLIR is capable of looking directly at a light source and still maintaining image quality, now I don't know the specifics but I do know that a real FLIR device has a krohmium lens to help filter out various things, what things I don't know, more like I wasn't allowed to know, but maybe you could get some more info about it through Raytheon, although I doubt it.

BTW the the image quality it sounds like you are going for is military quality..I doubt you'll get that with anything under $10k, especially at or 1000yrds.

[MichaelN]

BTW the the image quality it sounds like you are going for is military quality..I doubt you'll get that with anything under $10k, especially at or 1000yrds.

Even mil spec is generally not as good as the OP is hoping for. My brother works for a large scope manufacturer and said that reasonable thermal scopes are more like $20K...

[ScottH]

What wavelength are you interested in looking at? That's the first question.

Note that these camera do IR and visible, but they use two optical trains. You can design an achromatic system at two different wavelengths, but it is non-trivial.

You don't need DSP for this job and most focal planes that you can buy have a digital interface.

-Scott

Hello All, I am new and was recommended to this website by faculty at my university. I have a BS in chemical engineering and am currently working on the undergrad pre-reqs for a masters in EE. I would love to build a forward viewing IR scope. From what I understand this site sells DSP/microcontrolers with multipule analouge inputs with built in LCD screens. Do you sell microbolo meters here? Do you sell optics/lenses or know where I could order them or the scope case or where I could have a shell made.

The reason I need 2 inputs is because I want to display an IR image as well as the visible light image (overlayed on each other), there will be some challenges on the sensing device side to figure out how im going to do that without having redundant optics. Since IR is very close to the visible spectrum I am hoping I can use traditional visible light optics to focus IR along with the visible light without any/much distortion in the IR zooming/focusing, then the challenge would be to duplicate the image for 2 different sensing devices. Can the DSP/microcontrolers be battery operated?

I can program in VB but it would not be a stretch for me to learn C or C++ if nessicary, but VB would be ideal. I am also eventually going to take a DSP class and possibly a VLSI class in the future as well as some optics classes.

Thank you for the help, I am looking forward to the parallel journey into a masters degree and doing some cool things with electronics.

[ScottH]

Remember that the diffraction limit for longwave is proportionally bigger than in the visible. That's why the resolution is low: it's matched to the optics. High resolution means large optics. Getting the same resolution at 10 microns means you need an aperture that's 20 times bigger than in the visible.

I dont mind learning a new programming language but I do want to keep costs under control, typically pre-packaged IR devices run around 10k which is outragoiusly cost prohibitive when you can buy a microcontroler with LCD screen for 70-100$ and a microbolometer for 110$, the scope case and lenses might be a little more but not 10k more or anywhere close. I want to avoid buying something "out of the box" and having to mess with it, I am willing to take the time to build up the device so I get what I want.

I would need VERY high resolution, I will eventually want to make this into a long range scope where I can silouete a heat source image on top of a visible light image at maybe 1000 yards. I have seen visual outputs for various things on the market and the resolution is very poor and I know that much better can be attained, ever see the movie preditor and that was in the 80's.

Are there special forums to talk to the sales people at sparkfun to help me find the right parts I need. Thank you for the links for optics, I think the optics part may be pretty tough, but not impossible.

[MichaelN]

Remember that the diffraction limit for longwave is proportionally bigger than in the visible. That's why the resolution is low: it's matched to the optics. High resolution means large optics. Getting the same resolution at 10 microns means you need an aperture that's 20 times bigger than in the visible.

Indeed. Modern microbolometer sensors have pixels about 30 microns in size. Having a “high” resolution would mean a huge sensor – I don’t know of any company doing this.