Remote sensing
Data acquisition
Image processing
  Colour composites
  Geometric corrections
  Radiometric corrections
  Contrast enhancement
  Filtering
  Classification
  Visual interpretation
  Post-classification
  Indices
  Principal Component Analysis
  Combination of images
  Geospatial maps
  Combination of images and other data: DEM and DTM
Radar
GIS
 
Filtering
 
Filtering is an operation designed to improve images’ readability and/or to extract certain information from them. The principle of the various filters is to modify the numerical value of each pixel as a function of the neighbouring pixels’ values. For example, if the value of each pixel is replaced by the average of its value and those of its eight neighbours the image is smoothed, that is to say, the finer details disappear and the image appears fuzzier. In ‘mobile window filtering’ the programme computes for a given pixel the sum of the products of the neighbouring pixels’ values multiplied by coefficients given in a table of coefficients, then repeats the operation for the next pixel and so on.


For example, the filtered value of the pixel located at E5 is
(9*1/9) + (5*1/9) + (5*1/9) + (9*1/9) + (5*1/9) + (5*1/9) + (5*1/9) + (5*1/9) + (5*1/9) = 5.89, rounded up to 6.

This same mobile window has been applied to the three bands (red, green and blue) of the image on the left. The filtered result (to the right of the original) shows that many fine details (strands of hair, parts of the eyelashes, etc.) have disappeared and the image itself appears more blurred. This type of filtering, which eliminates small details, i.e., ones with high spatial frequencies, is also called ‘low pass filtering’.
The mobile window’s coefficients can also be adjusted to create filters that reveal specific directions in the image.

So, the values used in filter A enhance (in dark grey) the vertical boundaries seen in the image, whereas filter B enhances the horizontal ones. These effects are particularly noticeable at the edges of the eye’s iris.

 

Filters can also be used to reinforce the perception of an image’s sharpness. In the following example of a panchromatic (10m pixels) SPOT image of Brussels, the original image (to the left) was improved by applying a filter to enhance local contrast. The right-hand image clearly looks sharper, even though the processing did not actually add any information.