# Image Filters Panel

Filter image using different image filters.

Back to Index --> User Guide --> Panels

## Brief video demonstration

Use of filters in a brief video demonstration:

## Image Filter combo box.

Allows selection of a filter from a list of 2D and 3D image filters. Depending on the filter type some additional parameters should be specified in the HSize, Sigma, lambda, Type, Angle, Iter edit boxes.

List of available filters:

• Average, (2D) Matlab Averaging filter, see more in the Matlab documentation for fspecial and imfilter
• Disk, (2D) Matlab circular averaging filter (pillbox), see more in the Matlab documentation for fspecial and imfilter
• DNN Denoise, (2D) denoise images using deep neural network, available for Matlab R2017b and newer, requires Neural Network Toolbox and good GPU
• Gaussian, (2D) Matlab a rotationally symmetric Gaussian lowpass filter, see more in the Matlab documentation for fspecial and imfilter
• Gaussian, (3D) is based on Dirk-Jan Kroon implementation and uses the fact that a Gaussian kernel can be implemented as several 1D kernels.
• Frangi, (2D/3D), Hessian based Frangi Vesselness filter. This function uses the eigenvectors of the Hessian to compute the likeliness of an image region to contain vessels or other image ridges , according to the method described by Frangi 1998, 2001. Implementation is based on Hessian based Frangi Vesselness filter, written by Marc Schrijver and Dirk-Jan Kroon.
• Motion, (2D) Matlab filter to approximate the linear motion of a camera, see more in the Matlab documentation for fspecial and imfilter.
• Median, (2D) Matlab 2D median filter. Median filtering is a nonlinear operation often used in image processing to reduce "salt and pepper" noise. The median filter is more effective than convolution when the goal is to simultaneously reduce noise and preserve edges. See more in the Matlab documentation for medfilt2.
• Median, (3D) Matlab 3D median filter (Release 2017a and later). Median filtering is a nonlinear operation often used in image processing to reduce "salt and pepper" noise. The median filter is more effective than convolution when the goal is to simultaneously reduce noise and preserve edges. See more in the Matlab documentation for medfilt3. A short youtube demo
• Perona Malik anisotropic diffusion, (2D) - a filter written by Peter Kovesi to perform anisotropic diffusion of an image following Perona and Malik's algorithm. This process smoothes the regions while preserving, and enhancing the contrast at sharp intensity gradients.
• Unsharp, (2D) Matlab sharpens image using unsharp masking (imsharpen function, R2013a and above) or unsharpens contrast enhancement filter (fspecial and imfilter, R2012b and older).
• Wiener, (2D) Matlab 2D 2-D adaptive noise-removal filtering (wiener2 function). wiener2 lowpass-filters a grayscale image that has been degraded by constant power additive noise. wiener2 uses a pixel wise adaptive Wiener method based on statistics estimated from a local neighbourhood of each pixel.
• External: BMxD, (2D/3D) an optional filtering by block-matching and 3D collaborative algorithm. The filters are not supplied with MIB and should be intalled separately, please refer to the installation instruction in the System Requirements section. Reference: K. Dabov, A. Foi, V. Katkovnik, and K. Egiazarian, "Image Denoising by Sparse 3D Transform-Domain Collaborative Filtering," IEEE Transactions on Image Processing, vol. 16, no. 8, August, 2007. preprint at <http://www.cs.tut.fi/~foi/GCF-BM3D>. And M. Maggioni, V. Katkovnik, K. Egiazarian, A. Foi, "A Nonlocal Transform-Domain Filter for Volumetric Data Denoising and Reconstruction", IEEE Trans. Image Process., vol. 22, no. 1, pp. 119-133, January 2013. doi:10.1109/TIP.2012.2210725

Note! If HSize is specified with a single number then the size of the 3D Kernel is calculated based on pixel size of the dataset Menu->Dataset->Parameters. If HSize is specified with 2 numbers (i.e. 3;3) then the Kernel size is [3 x 3 x 3].

## Mode combobox

The Mode combobox allows to select part of the open dataset to apply the filters.

• 2D, shown slice, apply filter only for the currently shown slice
• 3D, current stack, apply filter only for the currently shown stack
• 4D, complete volume, apply filter for complete dataset

## Options combobox

The Options combobox allows to choose what to do with the dataset after filtration

• Apply filter, just filter the image and show result on the screen
• Apply and add to the image, filter the image and add the result to the original image
• Apply and subtract from the image, filter the image and subtract the result from the original image

## Various settings for the filters

Here is a set of edit boxes (3D, Type, HSize, lambda, Sigma, beta2, beta3) that define additional parameters for the filters. Depending on the selected filter one or more of these edit boxes may be disabled.

## The Filter button

Press this button to start the filtering.

Back to Index --> User Guide --> Panels