We introduce an image enhancement method referred to as the watershed-based
maximum-homogeneity filter. This method first uses watershed analysis to s
ubdivide the image into homogeneous pixel clusters called catchment basins.
Next, using an adaptive, local, catchment-basin selection scheme, similar
neighboring catchment basins are combined together to produce an enhanced i
mage. Because the method starts with watershed analysis, it can preserve ed
ge information and run with high computational efficiency. Illustrative res
ults show that the method performs well relative to other popular nonlinear
filters.