Features
Examples
Denoising Filters
Denoising filters remove unwanted high frequency signal typically by blurring an image.
(smoothing with intensities biased by distance)
(unbiased smoothing, structured neighborhood)
(edge preserving smoothing, structured neighborhood)
(edge preserving smoothing, noise estimation)
Background Subtraction
Background is unwanted signal typically from out-of-focus light. HIP provides the following filters:
(removes low frequency background which is estimated by a Gaussian smoothing)
Morphological Operators
Morphological operators are often used on binary images that represent the foreground of an image. However, HIP allows users to use any of the morphological operations on grayscale images as well. HIP also allows users to apply these filters iteratively creating a smoother effect (see discussion below).
(min filter)
(max filter)
(dilation followed by erosion)
(erosion followed by dilation)
Feature Detection
Feature detection filters identify areas with a particular quality in a local neighborhood. The resulting image emphasizes where those features are most predominate. For an example the Laplacian of Gaussian filter is a robust edge detector.
(edge detection using smooth second order derivative)
(uniformity or "texture" detection)
(edge detection using first order derivative)
(edge detection using second order derivative)
Combined Std and Var
(returns both the standard deviation and variance images)
Other Operations
Here is a list of other helpful filters: