IHS Transformation

See how to perform an IHS transformation in the SPRING


To describe an object color properties in an image, normally the human eye is not capable to distinguish the blue, green and red proportions present, and just evaluate the intensity (I), the hue (H) and the saturation (S).

The intensity or brightness is a total energy measure involved in all wave length, so it is responsible for the brightness sensation of this energy coming inside the eye.

The hue or object color is an average wave length of the light that reflects or if emitted, defining, thus, the object color.

The saturation or purity express the wave length interval around the average wave length, where the energy is reflected or transmitted. A high saturation value results in a spectrally pure color, and a low value indicates a mixture of wave length producing pale tones.

The IHS space can be graphically represented by a cone. The spatial relation between the RGB and IHS is presented in the figure below.

  • The distance from a point to the source or the cone top represents the intensity.
  • The radial distance from a point to the cone central axis represents the saturation.
  • The hue is represented as a radial sequence around the saturation circles and the intensity axis.

Because these three parameters are independent they can be analyzed and modified separately, for a better adjusting of colors to the visual system characteristics.

The components transformation red, green and blue (RGB) into the intensity, hue and saturation (IHS) components can be used to produce colored compositions with reduced correlation interband, consequently with a better utilization of the color space, and combine different image types or image from different sensors. These transformations are performed through mathematical algorithms that relates the RGB to the IHS space.

To produce colored compositions, three bands are selected from an image and each band is associated to an RGB component. The IHS "pixel" by "pixel" transformation is executed. Each "pixel" in the output image will have a corresponding point in the IHS space. The result is a set with three new images: one with intensity, one with hue and another one with saturation. These images are enhanced, such that the intensity and saturation interval are expanded using conventional enhanced contrast techniques, they are then transformed back to the RGB space, which gives a better separation of the desired aspects.

The IHS transformation can be used in the image combination of different sensors and spatial resolution, as an union of SPOT-HRV (pan-chromatic) and TM-Landsat images, for example.

The procedure involves computing the components H, I and S from the three selected bands of the TM and apply the contrast in the H and S resulting components, in the SPOT image. The I component is substituted in the SPOT image, and the inverse transformation (IHS-RGB) is applied.

After the transformation, the colored image will have a spatial resolution of the SPOT image and spectral resolution of the three TM bands.

In the SPRING, besides the IHS transformation and its inverse, the user can generate a synthetic image from the three components (R G, and B or I, H and S). This operation is also possible, using the Contrast enhanced interface.


See also:
About other Image Processing techniques.
Options Menu for Image Processing.