Passive IR temperature measurement vs active IR measurement

[stoyanoff]

Greatings! Could anyone explain to me what`s difference between measuring temperature with passive IR sensor and measuring with active sensor (by adding IR diode)? Does it make the measurement move accurate?!
Thanks!

 

[FvM]

I have no idea how you want to implement an "active temperature sensor" with an IR diode. What should be the measurement principle? Please clarify.

 

[stoyanoff]

I have an infrared sensor array. When I turn it against an hot object a few of it`s pixel change their temperature value according to the object and background temperature difference. But sometimes the test object has same temperature as the background. So read in Internet if I add a stress source(IR diode) I can use this technique for recognition of differend surfaces. So how could I achieve this? Do I have to add just a IR diode (I have a few TSHE5410) or I have to blink the diode with pwm signal? How far from the sensor could be situate the objects? I need to do this recognition up to 4 m.
Thanks!

 

[FvM]

I don't understand the meaning of "stress source". If the method makes any sense, it would use an IR "illimination" within the spectral sensitivity of the sensor. You didn't tell about it, but I guess it's > 7 µm far IR. A 900 nm IR diode won't be seen by the sensor.

 

[stoyanoff]

Could you offer me an IR diode with wavelenght between 2-20um? I`m not sure about the name of this type of IR diodes and I can`t find anything appropriate.
Thanks!

 

[jiripolivka]

I have an infrared sensor array. When I turn it against an hot object a few of it`s pixel change their temperature value according to the object and background temperature difference. But sometimes the test object has same temperature as the background. So read in Internet if I add a stress source(IR diode) I can use this technique for recognition of differend surfaces. So how could I achieve this? Do I have to add just a IR diode (I have a few TSHE5410) or I have to blink the diode with pwm signal? How far from the sensor could be situate the objects? I need to do this recognition up to 4 m.
Thanks!

The problem you have just described is a radiometry problem. By illuminating a target you will be able to detect target presence but not its temperature. You already know its temperature equals the ambient, so why to add the illumination?

Determining target temperature also depends on target albedo or emissivity. If you illuminate such target, the response will be driven by its emissivity but not linked to its (ambient) temperature.

You possibly can improve your system by adjusting background temperature, using a heated or cooled plate.

 

[stoyanoff]

I don`t need to measure the temperature of the object! Detecting a presence is OK! As I said the backgroun temperature and the object teperature becomes equal at noon when the sun heats the background. Now I`m using the difference between background and object temperature to detect presence but when the both temperatures become almost equal....

 

[jiripolivka]

I don`t need to measure the temperature of the object! Detecting a presence is OK! As I said the backgroun temperature and the object teperature becomes equal at noon when the sun heats the background. Now I`m using the difference between background and object temperature to detect presence but when the both temperatures become almost equal....

Okay, so please check again what I wrote. To detect an object with the radiative temperature equal to the ambient requires to change the contrast so the temperature difference allows to detect the object. The best way is to vary or change the temperature of the background, you can illuminate the object scene, so the reflected "light" can confirm the object presence in otherwise "void" pixel.
If you choose an illuminator, make sure its radiation spectral response covers your sensor sensitivity response. If you work in IR, if the LED or laser is not available at the particular spectral response of your sensor or camera, you can use a tungsten lamp, with or without a filter. If you scan the scene by a rotating mirror or lamp, you can improve the contrast by demodulating the target response by a synchronous detection of the output signal at the illumination frequency. Voids in the scene will not respond, so there are no objects in a void.

 

[FvM]

Could you offer me an IR diode with wavelenght between 2-20um?

You'll find a few mid IR leds covering the 2 to 4.5 µm wavelength range. They are probably expensive as most mid and far IR components. Thermal sources with a chopper are most likely achieving higher efficiency for broadband IR illumination.

My general thought about the problem is this: If you can't detect the object by it's passive radiation, you should use an active sensing method in a wavelength range that can be easily managed, e.g. near IR or VIS.