medfilt3

R2026b

3-D median filtering

Syntax

B = medfilt3(A)

B = medfilt3(A,[m n p])

B = medfilt3(___,padopt)

Description

B = medfilt3(A) filters the 3-D image A with a 3-by-3-by-3 median filter. By default, medfilt3 pads the image by replicating the values in a mirrored way at the borders.

B = medfilt3(A,[m n p]) performs median filtering of the 3-D image A in three dimensions. Each output voxel in B contains the median value in the m-by-n-by-p neighborhood around the corresponding voxel in A.

B = medfilt3(___,padopt) controls how medfilt3 pads the array boundaries.

Examples

Use Median Filtering to Remove Outliers in 3-D Data

Create a noisy 3-D surface.

[x,y,z,V] = flow(50);
noisyV = V + 0.1*double(rand(size(V))>0.95) - 0.1*double(rand(size(V))<0.05);

Apply median filtering.

filteredV = medfilt3(noisyV);
filteredVnopad = medfilt3(noisyV,[3 3 3],"zeros");

Display the noisy and filtered surfaces. Observe that filtered surfaces have lesser noise than the noisy surface. Also observe that when you specify zero padding, the edges of the filtered surface are rougher than the default symmetric padding.

hpatch1 = patch(isosurface(x,y,z,noisyV,0));
isonormals(x,y,z,noisyV,hpatch1)
set(hpatch1,FaceColor="red",EdgeColor="none")
daspect([1,4,4])
view([-65,20]) 
title("Noisy Surface")
axis tight off
camlight left
lighting gouraud

subplot(1,2,1)
hpatch2 = patch(isosurface(x,y,z,filteredV,0));
isonormals(x,y,z,filteredV,hpatch2)
set(hpatch2,FaceColor="red",EdgeColor="none")
daspect([1,4,4])
view([-65,20])
title("Symmetric Padding")
axis tight off
camlight left 
lighting gouraud

subplot(1,2,2)
hpatch3 = patch(isosurface(x,y,z,filteredVnopad,0));
isonormals(x,y,z,filteredVnopad,hpatch3)
set(hpatch3,FaceColor="red",EdgeColor="none")
daspect([1,4,4])
view([-65,20])
title("Zero Padding")
axis tight off
camlight left 
lighting gouraud

Input Arguments

A — Input image

3-D numeric array | 3-D logical array

Input image, specified as a 3-D numeric or logical array.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64 | logical

[m n p] — Neighborhood size

[3 3 3] (default) | 3-element vector

Neighborhood size, specified as a 3-element vector of positive odd integers.

Data Types: single | double | int8 | int16 | int32 | uint8 | uint16 | uint32

padopt — Padding option

"symmetric" (default) | "zeros" | "replicate"

Padding option, specified as one of the following values:

Value	Description	Example
"symmetric"	Pad array with mirror reflections of itself	$$\left[\begin{array}{ccc}3& 1& 4\\ 1& 5& 9\\ 2& 6& 5\end{array}\right]\to \left[\begin{array}{ccccccc}5& 1& 1& 5& 9& 9& 5\\ 1& 3& 3& 1& 4& 4& 1\\ 1& 3& 3& 1& 4& 4& 1\\ 5& 1& 1& 5& 9& 9& 5\\ 6& 2& 2& 6& 5& 5& 6\\ 6& 2& 2& 6& 5& 5& 6\\ 5& 1& 1& 5& 9& 9& 5\end{array}\right]$$
"replicate"	Pad array by repeating border elements	$$\left[\begin{array}{ccc}3& 1& 4\\ 1& 5& 9\\ 2& 6& 5\end{array}\right]\to \left[\begin{array}{ccccccc}3& 3& 3& 1& 4& 4& 4\\ 3& 3& 3& 1& 4& 4& 4\\ 3& 3& 3& 1& 4& 4& 4\\ 1& 1& 1& 5& 9& 9& 9\\ 2& 2& 2& 6& 5& 5& 5\\ 2& 2& 2& 6& 5& 5& 5\\ 2& 2& 2& 6& 5& 5& 5\end{array}\right]$$
"zeros"	Pad array with the value 0	$$\left[\begin{array}{ccc}3& 1& 4\\ 1& 5& 9\\ 2& 6& 5\end{array}\right]\to \left[\begin{array}{ccccccc}0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 3& 1& 4& 0& 0\\ 0& 0& 1& 5& 9& 0& 0\\ 0& 0& 2& 6& 5& 0& 0\\ 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0\end{array}\right]$$

Data Types: char | string

Output Arguments

B — Output image

3-D numeric array

Output image, returned as a 3-D numeric array of the same size and data type as the input image A.

Extended Capabilities

Thread-Based Environment

Run code in the background using MATLAB® backgroundPool or accelerate code with Parallel Computing Toolbox™ ThreadPool.

The medfilt3 function fully supports thread-based environments. For more information, see Run MATLAB Functions in Thread-Based Environment.

Version History

Introduced in R2016b

R2022b: Support for thread-based environments

medfilt3 now supports thread-based environments.

R2021b: Improved performance for small neighborhood sizes

The medfilt3 function shows improved performance for neighborhood sizes from [3, 3, 3] up to [31, 31, 31].

For example, this code is about 3x faster than in the previous release.

function timingTestMedfilt3
      load mristack;
      noisyV = imnoise(mristack,'salt & pepper',0.2);
      tic
      filteredV = medfilt3(noisyV);
      toc
end

The approximate execution times are:

R2021a: 0.24 s

R2021b: 0.08 s

The code was timed on a Windows^® 10, Intel^® Xeon^® Gold 5220 CPU @ 2.2 GHz test system by calling the function timingTestMedfilt3.