Products > Thermal Imaging

How to increase gain for old cores?

(1/3) > >>

Hi guys.
I found most of the old thermal imaging cores have a very low contrast when viewing low delta T scenes. Even the Seek have a higher contrast than most of them(although with more noise), is it possible to change the auto-gain limit higher?
Thank you.

You first need to detail the specific cores you have as then we will know what options are available to you  ;)


Short answer, probably no.

The limiting factor with uncooled microbolometer thermal imaging cores is noise, not gain. In a low delta-t scene the wanted signal (image) may be of the same magnitude - or even lower - than the sensor's noise. So if you were to increase the gain you'd just get the same signal-and-noise mix, just bigger.

You can 'force' gain in two ways: bigger (better) optics and / or temporal averaging. There's a practical limit to the aperture you can get in any optical system; I don't recall seeing any LWIR lenses of much more than f/0.7 (and those are rare). Temporal (frame) averaging will reduce the apparent noise, but at the expense of effective frame rate. All other things being equal, a rolling average of (say) 16 frames from the sensor will reduce the noise by a significant amount - but if anything moves in the images, you'll see sixteen ghosts of it as the movement works its way through the averaging buffer. (Many cameras already include frame-to-frame averaging to varying degrees as part of the standard noise reduction algorithms so you may find there's no benefit to additional averaging.

Greater detail can often be extracted from an image using a decrease in the temperature span to which the Palette LUT is applied. Visible Noise levels in the image increase as the span approaches single figures and this is normal. Careful selection of temperature span, Centre temperature and palette can provide optimum visible contrast in a low Delta T scene. In order to achieve such, the camera or core needs to be capable of manual span and centre temperature setting and some high contrast colour Palettes can highlight temperature transitions. High Contrast Rainbow is such a palette.

Finally, if a core offers high and low sensitivity modes of operation, the high sensitivity mode should be selected. This range usually covers approx -20C up to +120C. The low sensitivity mode provides greater temperature measurement capability but with higher noise levels present in the image and, in some cases, lower contrast in low Delta T scenes due to increased temperature span.


Bill W:
To add to things here, some of the early cores (Raytheon 2000AS and related) were quite limited in the available image processing to the point of being a fixed temperature - brightness map.  As said this was not so big an issue if noise limited anyway.

You might though have:
A sensor losing vacuum (early ULIS & Indigo)
A stuck iris (Raytheon BST)
Lost setup parameters (known with Raytheon BST)

The software and hardware routes to adjust a Raytheon BST gain and offset are in the forum.



[0] Message Index

[#] Next page

There was an error while thanking
Go to full version