14.5.2 Edge Enhancement
The brightness variations that define the edges and the text texture of objects in a
SAR scene are an important element in image interpretation. Objects of interest
may include those with a strong linear component, for instance geological fractures,
ocean wave fronts, or the edges of pack ice. An increase in local contrast may be
achieved by suppressing gradual brightness variations, which tend to obscure the
object of interest. Edges and linear features are composed of rapid brightness
variations, or high spatial frequencies, of a set of dark pixels next to a set of bright
pixels in an image. There a number of edge enhance algorithm which have been
designed to highlight these features. The examples presented below include the
high pass filter, the K-averaging filter and Fast Fourier Transform (FFT). The
variance and convex hull filters may be used for this purpose as well, but they are
better suited for reducing the appearance of speckle.
In the high pass, or Laplacian filtering process the low frequency component of
an image, as averaged over the filter window, is subtracted from the original image.
In order to avoid the occurrence of negative pixel values, a constant is added to the
intensity value of each pixel. An important consideration in the choice of the
filtering window is not only the size of the window, but also its shape. Square
Fig. 14.5 Synthetic color image
14 Digital Processing of SAR Data and Image Analysis Techniques
291
The brightness variations that define the edges and the text texture of objects in a
SAR scene are an important element in image interpretation. Objects of interest
may include those with a strong linear component, for instance geological fractures,
ocean wave fronts, or the edges of pack ice. An increase in local contrast may be
achieved by suppressing gradual brightness variations, which tend to obscure the
object of interest. Edges and linear features are composed of rapid brightness
variations, or high spatial frequencies, of a set of dark pixels next to a set of bright
pixels in an image. There a number of edge enhance algorithm which have been
designed to highlight these features. The examples presented below include the
high pass filter, the K-averaging filter and Fast Fourier Transform (FFT). The
variance and convex hull filters may be used for this purpose as well, but they are
better suited for reducing the appearance of speckle.
In the high pass, or Laplacian filtering process the low frequency component of
an image, as averaged over the filter window, is subtracted from the original image.
In order to avoid the occurrence of negative pixel values, a constant is added to the
intensity value of each pixel. An important consideration in the choice of the
filtering window is not only the size of the window, but also its shape. Square
Fig. 14.5 Synthetic color image
14 Digital Processing of SAR Data and Image Analysis Techniques
291