AutoIntensityClassification

Classifies all pixels of a grayscale image using the k-means clustering method.

Access to parameter description

As an introduction:

section Image Segmentation
section Unsupervised Classification

This algorithm segments the input image into a user-defined number of regions and assigns the same label to pixels sharing similar intensity characteristics. It implements the k-means clustering method, which is a standard for performing unsupervised classification into a known number of classes

$K$ .

Let

$S$ denote the set of image intensities. The input image can be a color image, so

$S$ is a set of vectors in which each element corresponds to an image channel.

$K$ vectors are randomly created to initialize the centers of the classes.
A subset $S'$ of $S$ is randomly selected to reduce the computation time.
The following process is iterated until minimization of the sum of the squared Euclidean distance of each vector in the cluster to its center (Within-Cluster Sum of Squares or WCSS):

All vectors of $S'$ are assigned to a class, according to the closest class centers, using the squared Euclidean distance.
The class centers are updated.

All vectors of the initial set $S$ are finally classified in the $K$ clusters to obtain the result classification.

A label is assigned to each pixel of the output image, corresponding to the class assigned to its input intensity vector.

$K$ is the number of classes to obtain, which is the main parameter of the algorithm. The size of the subset

$S'$ is also a parameter of the algorithm, given as a percentage of the input data.

Reference:
S. P. Lloyd, "Least square quantization in PCM", IEEE Transactions on Information Theory, vol. 28, no 2, pp. 129-137, 1982.

See also

Function Syntax

This function returns outputLabelImage.

// Function prototype
std::shared_ptr< iolink::ImageView >
autoIntensityClassification( std::shared_ptr< iolink::ImageView > inputImage,
                             int32_t classNumber,
                             int32_t dataPercentage,
                             std::shared_ptr< iolink::ImageView > outputLabelImage = NULL );

This function returns outputLabelImage.

// Function prototype.
auto_intensity_classification( input_image,
                               class_number = 2,
                               data_percentage = 100,
                               output_label_image = None )

This function returns outputLabelImage.

// Function prototype.
public static IOLink.ImageView
AutoIntensityClassification( IOLink.ImageView inputImage,
                             Int32 classNumber = 2,
                             Int32 dataPercentage = 100,
                             IOLink.ImageView outputLabelImage = null );

Class Syntax

// Command constructor.
AutoIntensityClassification();

/// Gets the inputImage parameter.
/// The input image.
std::shared_ptr< iolink::ImageView > inputImage() const;
/// Sets the inputImage parameter.
/// The input image.
void setInputImage( std::shared_ptr< iolink::ImageView > inputImage );

/// Gets the classNumber parameter.
/// The number of classes to detect (the label number of the output).
int32_t classNumber() const;
/// Sets the classNumber parameter.
/// The number of classes to detect (the label number of the output).
void setClassNumber( const int32_t& classNumber );

/// Gets the dataPercentage parameter.
/// The data percentage used for pre-computing the classification.
int32_t dataPercentage() const;
/// Sets the dataPercentage parameter.
/// The data percentage used for pre-computing the classification.
void setDataPercentage( const int32_t& dataPercentage );

/// Gets the outputLabelImage parameter.
/// The output label image where a one label represents one class.
std::shared_ptr< iolink::ImageView > outputLabelImage() const;
/// Sets the outputLabelImage parameter.
/// The output label image where a one label represents one class.
void setOutputLabelImage( std::shared_ptr< iolink::ImageView > outputLabelImage );


// Method to launch the command.
void execute();

# Property of the inputImage parameter.
AutoIntensityClassification.input_image
# Property of the classNumber parameter.
AutoIntensityClassification.class_number
# Property of the dataPercentage parameter.
AutoIntensityClassification.data_percentage
# Property of the outputLabelImage parameter.
AutoIntensityClassification.output_label_image

// Method to launch the command.
execute()

// Command constructor.
AutoIntensityClassification()

// Property of the inputImage parameter.
AutoIntensityClassification.inputImage
// Property of the classNumber parameter.
AutoIntensityClassification.classNumber
// Property of the dataPercentage parameter.
AutoIntensityClassification.dataPercentage
// Property of the outputLabelImage parameter.
AutoIntensityClassification.outputLabelImage

// Method to launch the command.
Execute()

Parameters

Class Name	AutoIntensityClassification

	Description	Type	Supported Values	Default Value
inputImage	The input image.	Image	Grayscale or Multispectral	nullptr
classNumber	The number of classes to detect (the label number of the output).	Int32	>=0	2
dataPercentage	The data percentage used for pre-computing the classification.	Int32	[1, 100]	100
outputLabelImage	The output label image where a one label represents one class.	Image		nullptr

Object Examples

auto foam = readVipImage( std::string( IMAGEDEVDATA_IMAGES_FOLDER ) + "foam.vip" );

AutoIntensityClassification autoIntensityClassificationAlgo;
autoIntensityClassificationAlgo.setInputImage( foam );
autoIntensityClassificationAlgo.setClassNumber( 2 );
autoIntensityClassificationAlgo.setDataPercentage( 100 );
autoIntensityClassificationAlgo.execute();

std::cout << "outputLabelImage:" << autoIntensityClassificationAlgo.outputLabelImage()->toString();

foam = imagedev.read_vip_image(imagedev_data.get_image_path("foam.vip"))

auto_intensity_classification_algo = imagedev.AutoIntensityClassification()
auto_intensity_classification_algo.input_image = foam
auto_intensity_classification_algo.class_number = 2
auto_intensity_classification_algo.data_percentage = 100
auto_intensity_classification_algo.execute()

print( "output_label_image:", str( auto_intensity_classification_algo.output_label_image ) )

ImageView foam = Data.ReadVipImage( @"Data/images/foam.vip" );

AutoIntensityClassification autoIntensityClassificationAlgo = new AutoIntensityClassification
{
    inputImage = foam,
    classNumber = 2,
    dataPercentage = 100
};
autoIntensityClassificationAlgo.Execute();

Console.WriteLine( "outputLabelImage:" + autoIntensityClassificationAlgo.outputLabelImage.ToString() );

Function Examples

auto foam = readVipImage( std::string( IMAGEDEVDATA_IMAGES_FOLDER ) + "foam.vip" );

auto result = autoIntensityClassification( foam, 2, 100 );

std::cout << "outputLabelImage:" << result->toString();

foam = imagedev.read_vip_image(imagedev_data.get_image_path("foam.vip"))

result = imagedev.auto_intensity_classification( foam, 2, 100 )

print( "output_label_image:", str( result ) )

ImageView foam = Data.ReadVipImage( @"Data/images/foam.vip" );

IOLink.ImageView result = Processing.AutoIntensityClassification( foam, 2, 100 );

Console.WriteLine( "outputLabelImage:" + result.ToString() );

AutoIntensityClassification

- Function Syntax

+ Class Syntax

Parameters

Object Examples

Function Examples

Function Syntax

Class Syntax