Figure 7.2 The histogram of an image is a plot of pixel value versus the number of pixels having each value. It provides insight as to what range of pixel values holds the most useful information. Although pixel values here range from 130 to 11265, values are strongly concentrated near that of the background sky.
CCD camera, for example, has a characteristic spike at a low pixel value and virtually no pixels outside the spike. In contrast, a dark frame shows a narrow spike but with an exponentially decaying tail of high-value pixels. The spike comes from the vast majority population of well-behaved normal pixels, while the tail reflects the presence of a minority population of hot pixels.
Healthy flat-field frames usually show a broad peak centered on the average value of the flat field—but a sharp cutoff on the high side reveals saturation, a pathological condition for a flat-field frame.
With astronomical images, of course, you see a broad spike from sky pixels, and you can tell whether the values cluster at the low end (underexposure) or fill a reasonable portion of the dynamic range of the CCD. Pixels that are parts of star images form a tail that declines exponentially toward high values. Humps and bumps superimposed on the exponential tail are generally from the pixels in galaxies and nebulae. A sharp cutoff at high pixel values marks the onset of saturation in the CCD chip, usually affecting only the cores of bright star images.
Planetary images often display bimodal distributions with one peak from the sky surrounding the planet, and a second peak at higher pixel values from its disk.
Mars often shows a spike at the high end of its peak from the high-value pixels in the polar cap.
Was this article helpful?