Component: SegmentSeeded

SegmentSeeded

Segments a directory of images with various methods, using a seed (a previous mask helping in refining the segmentation). The method choices are:

Otsu's method thresholds the object intensities so that the intra-class variance is maximized. The calculated threshold value can be multiplied with a correction constant, to improve the result.

Shape method thresholds the objects at several intensity levels and searches for objects fitting the size and roundness criteria given in the parameters.

ConstantArea method thresholds the objects at several intensity levels and returns a mask where a certain ratio of object area is included in the mask. Filling/size restrictions are applied after masking. Use constant to control the area ratio. A minimization function is used, so the area fraction may not be accurate.

KNN method takes a CSV file with cluster numbers and color values, and seeks the nearest neighbor for each pixel.

kernel method takes a CSV file with cluster numbers and color values, and seeks the nearest neighbor for each pixel by using any metric provided in the 'kernel' parameter. NOTE: Since custom metric is used, the full distance matrix from each pixel to color value is calculated. It easily consumes a lot of memory if either is large.

quadratic method takes a CSV file with cluster numbers and color values, and uses a quadratic classifer to get pixel classes.

Version	1.0
Bundle	anima
Categories	Image Processing
Authors	Ville Rantanen (ville.rantanen@helsinki.fi)
Issue tracker	View/Report issues
Requires	Matlab
Source files	component.xml seededshapesegment.m seededareasegment.m seededotsusegment.m knnsegment.m imagesegment.m execute.m
Usage	Example with default values SegmentSeeded( in = ImageList, //seed = ImageList, //colors = CSV, clearBorders = true, constant = 0.5, corr = 1, fill = true, kernel = "mean(exp(-x))", maxArea = 0, maxEcc = 1, maxRound = 1, method = "Otsu", minArea = 0, minEcc = 0, minInt = 0, minRound = 0, numbers = false )

Version

1.0

Bundle

anima

Categories

Image Processing

Authors

Ville Rantanen (ville.rantanen@helsinki.fi)

Issue tracker

View/Report issues

Requires

Matlab

Source files

component.xml seededshapesegment.m seededareasegment.m seededotsusegment.m knnsegment.m imagesegment.m execute.m

Usage

Example with default values

Inputs

Name	Type	Mandatory	Description
in	ImageList	Mandatory	A directory with grayscale image files. (or possibly RGB for 'KNN')
seed	ImageList	Optional	Directory of mask files that restrict the pixels used for segmentation.
colors	CSV	Optional	CSV file with 2 or 4 columns. First with cluster number, and other(s) with color values (range 0-1), order R G B.

Name

Type

Mandatory

Description

ImageList

Mandatory

A directory with grayscale image files. (or possibly RGB for 'KNN')

seed

ImageList

Optional

Directory of mask files that restrict the pixels used for segmentation.

colors

CSV

Optional

CSV file with 2 or 4 columns. First with cluster number, and other(s) with color values (range 0-1), order R G B.

Outputs

Name	Type	Description
mask	ImageList	BW PNG images.
perimeter	ImageList	BW PNG images of the perimeter of the mask.
numbered	ImageList	BW PNG images of the numbers of the objects in the mask. Created only if parameter numbering is set true.

Name

Type

Description

mask

ImageList

BW PNG images.

perimeter

ImageList

BW PNG images of the perimeter of the mask.

numbered

ImageList

BW PNG images of the numbers of the objects in the mask. Created only if parameter numbering is set true.

Parameters

Name	Type	Default	Description
clearBorders	boolean	true	Clear objects touching the image edges. Applicable to all methods.
constant	float	0.5	The area ratio in ConstantArea method.
corr	float	1	Correction multiplier for threshold value in 'Otsu' method.
fill	boolean	true	Fill holes in segmented objects. Applicable to all methods.
kernel	string	"mean(exp(-x))"	Kernel in 'kernel' method. Cluster will be chosen by maximizing the kernel value. The 'x' is the vector of distances for the colors in the cluster.
maxArea	float	0	Maximum accepted pixel area in all methods. Use 0 for no limit.
maxEcc	float	1	Maximum accepted eccentricity value in 'Shape' method.
maxRound	float	1	Maximum accepted roundness value in 'Shape' method.
method	string	"Otsu"	Method for image segmentation. Choose from 'Otsu', 'Shape', 'ConstantArea','KNN','Kernel','Quadratic'.
minArea	float	0	Minimum accepted pixel area in all methods.
minEcc	float	0	Minimum accepted eccentricity value in 'Shape' method.
minInt	float	0	Minimum accepted intensity value in 'Shape' and 'Otsu' methods. Ranges 0-1.
minRound	float	0	Minimum accepted roundness value in 'Shape' method.
numbers	boolean	false	Draw images of the numbers of the objects in the mask. This can be a lengthy process if there are huge amounts of objects.

Name

Type

Default

Description

clearBorders

boolean

true

Clear objects touching the image edges. Applicable to all methods.

constant

float

0.5

The area ratio in ConstantArea method.

corr

float

Correction multiplier for threshold value in 'Otsu' method.

fill

boolean

true

Fill holes in segmented objects. Applicable to all methods.

kernel

string

"mean(exp(-x))"

Kernel in 'kernel' method. Cluster will be chosen by maximizing the kernel value. The 'x' is the vector of distances for the colors in the cluster.

maxArea

float

Maximum accepted pixel area in all methods. Use 0 for no limit.

maxEcc

float

Maximum accepted eccentricity value in 'Shape' method.

maxRound

float

Maximum accepted roundness value in 'Shape' method.

method

string

"Otsu"

Method for image segmentation. Choose from 'Otsu', 'Shape', 'ConstantArea','KNN','Kernel','Quadratic'.

minArea

float

Minimum accepted pixel area in all methods.

minEcc

float

Minimum accepted eccentricity value in 'Shape' method.

minInt

float

Minimum accepted intensity value in 'Shape' and 'Otsu' methods. Ranges 0-1.

minRound

float

Minimum accepted roundness value in 'Shape' method.

numbers

boolean

false

Draw images of the numbers of the objects in the mask. This can be a lengthy process if there are huge amounts of objects.

Test cases

Test case	Parameters▼	IN in	IN seed	IN colors	OUT mask	OUT perimeter	OUT numbered
case1_otsu	properties	in	seed	(missing)	mask	perimeter	numbered
	method=otsu, minArea=3, maxArea=0, corr=1.3, numbers=true, metadata.timeout=60
case2_shape	properties	in	seed	(missing)	mask	perimeter	(missing)
	method=shape, constant=0.333, minArea=5, maxArea=300, minEcc=0.0, maxEcc=0.75, minRound=0.6, minInt=0.15

Test case

Parameters▼

IN
in

IN
seed

IN
colors

OUT
mask

OUT
perimeter

OUT
numbered

case1_otsu

properties

seed

(missing)