Most satellite images are very dark and virtually nothing of interest can be deciphered in them.  The reason for such a low contrast is that many natural surface features have a low range of reflectances in any waveband.  An 8-bit detector which can record 256 gray levels must be able to record a reflectance value for the brightest snow (say a value of 255) and also for the darkest of rocks like black basalt (say a value of 0).  If these sensors were not designed to record these values they might not be able to record information of these surface types.  An average scene however does not call for this entire range of reflectances (ranging from 0-255) and the DNs are often compressed into a small part of the available range. 

Linear contrast stretching involves altering the distribution and range of DN values in an image.  It entails translation of the image pixel values from the observed range Vmin to Vmax to the full range of the display device (generally 0-255, with an 8-bit image).  V is the pixel value observed in the image with Vmin being the lowest value and Vmax the highest.  The pixel values are scaled so that Vmin maps to the value 0 and Vmax maps to 255.  Intermediate values change correspondingly, but retain their relative positions, so that the values in the middle of the range map to 127.  Even if the middle range of the pixel value is calculated to be 127.5, it cannot be mapped to that value because the computer display system can handle only integers (1, 2, 3255).

The image in Fig 1 has DN values that range from 80 to 144.  A linear contrast stretch assigns new DN values to output image by assigning to the lowest and highest DN in the input image  values of 0 and 255 respectively in the output image and stretching all the intervening digital numbers accordingly.  The output image is shown in Fig 2.  A simple formula used to determine the digital numbers in the stretched image is:

Where DNmin and DNmax are the minimum and maximum DN values in the image and DNst is the new DN value of the stretched image.

Fig 1:  Unstretched image of the area around Aligarh town, with DN values ranging from 80 to 144

Fig 2:  Linear contrast stretched image of the same area as in Fig 1.


Contrast stretching, which involves altering the distribution and range of DN values, is usually the first and most important step applied to image enhancement. A cursory look at any image will reveal that modifying the range of light and dark tones (gray levels) is often the single most important information revealing operation performed on the scene. Contrast stretching an image by computer processing of digital data (DNs) is a common operation, although we need some user skill in selecting specific techniques and parameters (range limits).


The human vision is more adept at discerning colours than distinguishing shades of gray.  A digital image processing system allows for assigning a different colour to each DN or a range of DNs in an image.  This makes it possible to assign different colours to different groups of pixels in a single-band image even though normally such an image would be displayed in shades of gray.  Such a process whereby different groups of pixels having DN values lying between certain limits, is called density slicing.  A density sliced surface is therefore more readable when it comes to distinguishing between different land cover types.

The number of slices and the range of DNs to be assigned to each slice are determined interactively by the user, and depends on the particular scene and the kind of information that is required to be extracted.  It depends on the range of DN values a particular feature may have. A density sliced version of the single band gray scale image shown in Figs 1 and 2 is displayed in Fig 3.

Fig 3:  Density sliced image of the same dataset as in Figs 1 and 2.

 A disadvantage of the density sliced image is that subtle detail is lost because a range of DNs is assigned a single colour.

Notes & Handouts

The Himalayas

Kumaon Himalayas

Askot Basemetals



This website is hosted by

S. Farooq

Department of Geology

Aligarh Muslim University, Aligarh - 202 002 (India)

Phone: 91-571-2721150