SelectComponents

✖
`SelectComponents`

✖

SelectComponents[{image,lmat},crit]

selects components of image indicated by the label matrix lmat that satisfy crit, replacing other parts with black.

✖

SelectComponents[image,crit]

selects connected components of image.

✖

SelectComponents[…,"prop",n]

computes the property "prop" and selects the first n in sorted order.

✖

SelectComponents[…,"prop",n,p]

sorts computed properties using the ordering function p.

Details and Options

SelectComponents can be used to select image components with specific desired characteristics. Selection may be performed on components' location, shapes, and intensity properties.
Component selection is also used to filter undesired image parts, including small, large, border touching, or specifically shaped components.
SelectComponents works with 2D or 3D images and label matrices.
A label matrix lmat is an array of non-negative integers, in which each integer represents a component, and 0 represents the background.
If the label matrix lmat is not given, it is computed using the connectivity of nonzero pixels in image.
SelectComponents selects components based on the given criteria crit. Components that are not selected are replaced with 0 (black) in both image and lmat.
Possible forms of crit include:
Large components bigger than a globally determined threshold

Small components smaller than a globally determined threshold

f components that satisfy f
When a pure function f is used, component properties can be accessed using #prop₁, #prop₂, …. A component is selected if f returns True.
For the complete list of possible properties, see the reference page for ComponentMeasurements.
SelectComponents[{image,lmat},…] returns a modified version of {image,lmat} where unselected components are replaced with 0.
SelectComponents[lmat,…] can also be used to select components from a label matrix lmat.
SelectComponents[…,"prop",n] computes "prop" for all components and returns the first n, sorted using Sort.
SelectComponents[…,{"prop₁","prop₂",…},n] computes and sorts multiple properties.
SelectComponents[…,"prop",-n] returns the last n components.
SelectComponents takes a CornerNeighbors option. The default setting is CornerNeighbors->True.

Examples

open allclose all

Basic Examples (3)Summary of the most common use cases

Select the outermost components in an image:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-uv45bi

Out[1]=1

Select three smallest components:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-e5o5qv

Out[1]=1

Select 3D image components:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-347vk

Out[1]=1

Scope (7)Survey of the scope of standard use cases

Use a single property to select components:

In[2]:=2

✖

https://wolfram.com/xid/0dqtl9sfuhf-hzyjni

Out[2]=2

Use multiple properties:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-bu968j

Out[1]=1

Select large components from a binary image:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-ef9njm

Out[1]=1

Select small components:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-pgkyn1

Out[1]=1

Select the six smallest components:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-osq8qd

Out[1]=1

Select the six largest components:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-c82tpf

Out[1]=1

Select components of a label matrix:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-739xv0

Options (1)Common values & functionality for each option

CornerNeighbors (1)

By default, SelectComponents considers diagonally adjacent corners to be connected:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-ys0ns

Out[1]=1

Use CornerNeighborsFalse to assume otherwise:

In[2]:=2

✖

https://wolfram.com/xid/0dqtl9sfuhf-636qtk

Out[2]=2

Applications (5)Sample problems that can be solved with this function

Select all the border components:

In[4]:=4

✖

https://wolfram.com/xid/0dqtl9sfuhf-7i90cj

In[5]:=5

✖

https://wolfram.com/xid/0dqtl9sfuhf-j49zv9

Out[5]=5

Select all interior components:

In[7]:=7

✖

https://wolfram.com/xid/0dqtl9sfuhf-54a7cr

Out[7]=7

Select the four roundest objects:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-cucx8n

Out[1]=1

Find objects smaller than 500 pixels that have holes and do not touch an image border:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-l3ii5

In[2]:=2

✖

https://wolfram.com/xid/0dqtl9sfuhf-9lw2l

Out[2]=2

Extract segments belonging to numbers in an image of a calendar:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-y5y67

Out[1]=1

Find the river in a satellite image by selecting the longest component:

In[1]:=1

✖

https://wolfram.com/xid/0dqtl9sfuhf-ev5h3c

Out[1]=1

Top