HOW TO FIND THE TEMPERATURE VALUE IN THE THERMAL CAMERA IMAGE?

I AM TRYING TO WRITE A SOFTWARE THAT DETECTS THE TEMPERATURE VALUES IN THE IMAGE IN DEGREES WHEN A COMPUTER MOUSE LOOKS OVER THE THERMAL CAMERA IMAGE. HOW CAN I WRITE THIS?

[Özgür Uzunkaya]

degrees should be readable when mouse hovers over the image

[dpb]

Grab the data value at the location and use the definition of the image correlation to temperature, maybe?

One would presume the vendor has algorithms for this you can use...

[Özgür Uzunkaya]

This is my project. I have to write before I buy. Can you help me?

[dpb]

Nope, not really, never used one other than with a vendor-supplied toolset, and that was only a one-time exercise at a power plant monitoring/finding hotspots on the boiler.

Bth, here's bound to be documentation for whatever product you're using -- I don't know if there's an industry standard that's followed or not.

Research is part of a project scope, not passing off the work to somebody else...

Answer 1

[Walter Roberson]

See https://www.mathworks.com/matlabcentral/answers/1723490-how-do-i-map-grayscale-values-to-get-temperature#comment_2169050

Answer 2

[Image Analyst]

See attached demo. Adapt as needed to your image.

[Özgür Uzunkaya]

What do the values ​​on the color scale mean? The values ​​I know should be as I share. How can I change the values ​​of the numbers here?

[Walter Roberson]

That colorbar is labeled in nm, nanometers.

400 nm is "violet", and the lower limit of visible is about 380 nm, which is about where the colorbar ends. 400 nm is the limit of UV-A, "long wavelength ultraviolet"

700 nm is middle red; the upper limit of visible is about 750 nm. 700 nm is the limit of the IR-A band (near infrared), 3864 K. If I read charts correctly, 400 nm should be about twice that temperature.

[Image Analyst]

The units of a thermal image should be degreed Celsius, not nanometers. It looks like someone just assigned text to the tickmarks that roughly corresponded to the colors of the rainbow/visual spectrum rather than temperature. Not sure why they did that. It might be applicable if you wanted to give the dominant wavelength in a true color image, but it would certainly be deceptive if that colorbar with those tick marks was displayed next to a pseudocolored thermal image. Look at my original screenshot image thermal_image.png and you can see that image gives the colorbar in units of temperature, not wavelength.

[Özgür Uzunkaya]

I don't fully understand if the values ​​in this color bar are in degrees Celsius, is it 50 degrees Celsius red, 200 degrees Celsius dark blue?

[Walter Roberson]

No, that is an uncalibrated color bar showing the relationship between array values and colors. No temperature calibration is implied by that image.

We cannot conclude anything about temperature. By convention, lowest data values would normally be at the bottom of the color bar, but obviously not in this case. Is it possible that blue corresponds to hotter? It is not impossible.

If you look at the DIAS example that Image Analyst posted then it shows the same colors and blue at the bottom is associated with the lowest temperature, so the only calibrated example we have here would have blue be cooler and red be hotter. But if you remember physics, "red hot" is the coolest temperature that emits visible light, and the hottest that emits visible light is far blue. Hotter is bluer, ultraviolet is high energy in physics. So we cannot rule out blue being associated with hotter... but neither can we confirm it.