This problem is best explained with an example: Our image consists of 12 bit data with a minimum intensity of 78 and a maximum of 1630 counts out of 4095 possible. The distribution of intensities is shown in the histogram.
The on-screen appearance of digital images, that is, the way the data are displayed, is defined by the display palette or look-up table. It relates the different greyscales or colours to the range of pixel intensities that occur in the image. If the example is displayed with a linear greyscale palette where 0 counts are assigned black and 4095 counts assigned white, as in the graph below, the result may be an image that is very dark and low in contrast as shown in the example.
Most image processing programs offer an autoscale function which adjusts the greyscale palette so that the darkest pixel of an image is set to black and brightest to white. The contrast of the image display can further be optimised by changing the slope of the gradient or any background intensity may be blanked out by shifting the black threshold upwards.
Monochrome images can also be displayed in pseudo-colours, for example in order to appear as observed through the eyepiece of a microscope. Instead of a black and white palette, for example a black and green palette can be used:
Often false colour palettes are used because of the increased contrast they offer. This is due to the ability of the human eye to better distinguish colours than greyscales. A popular such palette is the black-blue-green-red pseudo-rainbow. Note that with the same slope of the gradient as above, the weaker details of the image are better visible:
