Infrared and NIR Camera Sensitivity

[bitsnbytes]

Problem: I'm trying to optimize an inexpensive digital camera with a CCD sensor for NIR video. As of right now (after removing the IR filter) it doesn't work well in the dark at all (even with a basic visible light filter ... obviously). I have to direct at least a little bit of regular light or using the remote control (IR LEDs) to illuminate an object in complete darkness to see it.

What I'd like to do now is figure out how much IR energy I need to direct at an object for a given distance to be able to make out what it is?

Second part is to figure out camera sensitivity (I figure its between 780nm and 1200nm which is why heat won't show up (unless I heat a knife glowing orange wait for it to cool just enough to the point that you can't see it in the dark and I get my white spot indicating heat source)) but how can I tell with certainty that it is between x um and y um?

I figure camera sensitivity ultimately defines IR intensity required.

My goal is to be able to clearly see outlines of people and vehicles in a basic surveillance application. (I cannot afford a camera with a microbolometer image sensor right now.)

I have seen IR Illuminators that go 300meters, but I need to hit at least 1,000ft - 2,500ft (size and weight is a major issue)...and again I need to figure out how much energy needs to be concentrated in a given area at anyone time (distance) to have my minimum usable quality of video (camera dependent).

Most important question, how much IR energy is considered lethal? People will be unknowingly exposed to IR during operation.

Any help and guidance to theory behind your answer would be greatly appreciated!

The knowledge will be used towards a project focused on developing technology to assist a non-profit criminal profiling agency in its work to recover abducted children.

This is by no means my area of expertise!

 

[Andy Resnick]

There is a lot in your post that gives me pause ('basic surveillance', for example), but it is true that silicon-based detectors found in commercial cameras etc. can be used to image NIR.

Bare silicon is sensitive out to about 1.1 um, but don't forget color cameras have a Beyer filter overlaid (in addition to the IR cut filter you removed). The Beyer filter will probably transmit some NIR, but I couldn't say what the details are. One simple check is to image a TV remote- the diode operates around 800 nm, IIRC.

In any case, the NIR spectrum is still dominated by sunlight- people emit around 10 um and are best imaged by a LWIR camera using HgCdTe (or something similar like PtSi) focal plane arrays. You will still be imaging light that reflects off an object, not the thermal signatures of targets.

Again, I almost reported this post. The comment 'Most important question, how much IR energy is considered lethal? People will be unknowingly exposed to IR during operation' tells me you don't know what you are doing.

These sites may help you:
http://www.naturfotograf.com/UV_IR_rev00.html
http://www.beyondvisible.com/

 

[Tea Jay]

I think there was an app to convert your cell phone camera, but frankly, your post is kind of scary, as in, it sounds fishy/doesn't add up.

 

[bitsnbytes]

Alright, let me try to put to rest any uncertainties:

1) This is not my area of expertise, I'm an FAA licensed pilot and studying computer science with an emphasis on software engineering while dabbling in sub-system design. Designing a NIR/FLIR imaging system is a specialty area in itself. So no, I don't know everything there is to do with the electromagnetic spectrum (as it applies to this project), specifically infrared. If I did, I wouldn't be here asking for help (I understand the concern though). No one else on the team has any experience with it either, so I got stuck with it. I started to explore this area to find an alternative to the $10k+ cameras required for FLIR (at the level my project requires). I began looking into this subject to try and figure out how much the system would weigh and asses its power consumption which I need to know for other parts of the project I'm working on. During my research, my interest peaked and now I'd want to build something for fun as a learning experience. (video/imaging sub-system)

2) This sounds fishy, what are you doing? My team and I are designing a UAV and this subject came up for its video/imaging sub-system. The UAV is being developed for a non-profit agency that assists law enforcement with criminal profiling, missing children and just about anything else you could think of. This doubles up as a project to facilitate experience and learning in respective fields while creating a product that can be used by law enforcement. If we end up modifying equipment to develop an 'in-house' NIR solution (until we can afford to buy FLIR equipment) we will go through the steps to get it licensed and approved by the appropriate government agency(ies) before using it in the general public. If it even makes it out of the testing site with the nightmare it will be to get a COA from the FAA.

I asked the question about lethal exposure to IR because I'm going to be playing around with it and I don't want to hurt myself in the process let alone anyone that happens to be in the area, and also to see if its even practical to use IR LED illuminators to 'see' targets at the given 1,000 - 2,500 foot distance. As in so much IR energy would be needed to allow the cameras to see that far, that anyone near it could be exposed to a 'lethal' (now that I think about it,that's probably not the best choice of words) amount of IR energy.

This is an area that interests me in a self-study capacity, which is why I asked for direction to the theory behind answers to specific questions.

The main thing I wanted to get from this post was: how to measure camera sensitivity and how to determine the amount of IR energy to see objects at the given distance. Then guidance to reading on the theory behind those answers so I could understand it, rather than only run with the answer.

Thanks.

 

[Tea Jay]

You can buy a FLIR for a couple grand now, and Fluke versions for even less, and you'd be able to pick up heat signatures. Used ones are on ebay from time to time for example.

With a true IR camera, you would not need to illuminate your targets, as their radiated heat would self-illuminate them for you...they would glow in the view finder, etc. This would also reduce the detectability of your device, as it will not emit a light that could be used to target it. (Passive collection)

You can typically adjust the camera settings for absorbance and transmittal characteristics optimized to resolve your targets, better for live feed, or set it to gather raw data that you process with software after the run, etc.

If making a flying surveillance device, the weight of the batteries, etc, to power a lighting system would be counter productive, and best avoided.


I'd like to add that an IR light is just light, and that in the context of providing illumination, it would be about as dangerous as a flashlight. A laser, or other concentrated beam of course is different.

I am a Certified Thermographer for example, with training in this field. I mostly use it to diagnose building problems though. I also do emergency and disaster response, and this is the aspect that overlaps your objective - in that finding people, even when they are hidden, is a priority.