Why does intensity graph prefer higher values?

[psychomanu]

Hi,

We use intensity plots to display relatively large images (2Kx2K), which are quite noisy in some parts, and less in other parts. The problem is that the apparent average intensity in the noisy regions is too high. It is a display problem, because if you zoom in it dissappears.

Just have a look at the included picture that displays noisy data (randoms between 0 and 1). As long as the number of pixels is low (relative to the number of pixels on the screen) there is no problem (left image). When displaying larger arrays you expect to see about the average value (0,5). Instead it prefers to display the brighter values. This is wrong behaviour IMHO. When an array has to be downsampled to be displayed on screen, it should happen in an unbiassed way, not preferring higher or lower vvalues. The IMAQ intensity plot behaves correctly, as does any image processing software I tested.

Any comments or solutions?

Thanks,

Manu

[psychomanu]

Hi again,

I posted the same question on the NI discussion forum ( http://forums.ni.com/ni/board/message?boar...ssage.id=136822 ) and they acknowledge that this is indeed wrong behaviour.

So be careful when reading values from intensity graphs of large arrays!

Manu