Image posterization from k-means results algorithmSee full documentation

image =	kMeansPosterizeImg (inImg,classImg,clustersCenters)
image =	kMeansPosterizeImg (inImg,classImg,maskImg,clustersCenters)

Detailed Description

Posterize an image using the results from k-means algorithm.

Fills the output image (that must have the same geometry as the input one), by replacing each pixel value of the input image ( ) with the value of the center of the cluster which it belongs to. Values of clusters centers are defined by attribute, and input image clustering is defined by attribute. These 2 attributes are usually initialized using K-means algorithm.

Here is an example of posterization algorithm applied to Lena RGB image and to results of K-Means algorithm application (4 clusters expected):

It is also possible to provide a mask image. In this case, the posterized image will have a value of 0 where the mask equals False. Here is an example of masked posterization:

Example of Python code :

Example imports

import PyIPSDK
import PyIPSDK.IPSDKIPLClassification as classif

Code Example

    # load input data from files
    inImg = PyIPSDK.loadTiffImageFile(inputImgPath)
    inClustersCenters, bReadRes = PyIPSDK.readFromXmlFile(inClustersCentersPath)
    inClassImg = PyIPSDK.loadTiffImageFile(inClassImgPath)
    
    # apply posterization algorithm, and get posterized image
    autoInitializedOutImg = classif.kMeansPosterizeImg(inImg, inClassImg, inClustersCenters)

Example of C++ code :

Example informations

Header file

#include <IPSDKIPL/IPSDKIPLClassification/Processor/KMeansPosterizeImg/KMeansPosterizeImg.h>

Code Example

    // load input image from file
    ImagePtr pInImg = loadTiffImageFile(inputImgPath);

    // load input class image from file
    ImagePtr pInClassImg = loadTiffImageFile(inClassImgFilePath);

    // load clusters centers
    ClustersCentersPtr pInClustersCenters = boost::make_shared<ClustersCenters>();
    readFromXmlFile(inCentersFilePath, *pInClustersCenters);
    

    // apply posterization algorithm on input data, and get posterized image
    ImagePtr pOutImg = classif::kMeansPosterizeImg(pInImg, pInClassImg, pInClustersCenters);