Kernel Density Estimator 2d

module demonstrating kernel density estimation in case of a 2d image More...

module demonstrating kernel density estimation in case of a 2d image

Overview

This application computes histogram of an input 2d image using a classical histogram algorithm. A second computation is then proceeded using a kernel density estimation method. Input image is loaded from a given input TIFF file. Histogram results are saved in a given csv file.

See also

histogram 2d algorithm

kernel density estimator 2d algorithm

Usage

This is a standalone script file which takes no input argument.

Here is a snapshot of input image used by the application and of corresponding results :

Source code documentation

We start by importing all necessary libraries:

import os
import sys, getopt
import PyIPSDK
import PyIPSDK.IPSDKIPLGlobalMeasure as glbmsr

Then we define the input parameters.

# define input image path
imagesSamplePath = PyIPSDK.getIPSDKDirectory(PyIPSDK.eInternalDirectory.eID_Images)
inputImgPath = os.path.join(imagesSamplePath,  "Lena_510x509_UInt8.tif")

# define output data path
outputDataPath = PyIPSDK.getIPSDKDefaultDirectory(PyIPSDK.eDefaultExternalDirectory.eDED_Tmp)
refCsvPath = os.path.join(outputDataPath,  "ref_histogram.csv")
estimCsvPath = os.path.join(outputDataPath,  "estim_histogram.csv")

We load our input image from the associated Tiff file, by calling the function ipsdk::image::file::loadTiffImageFile.

# opening of input image
inImg = PyIPSDK.loadTiffImageFile(inputImgPath)
nbTotPixels = inImg.getGeometry().getNbPixels()

We then apply the histogram computation on the input image using a classical histogram algorithm and a faster one based on a kernel density estimation algorithm.

# definition of histogram measure parameters
binMin = 0
binMax = 255
binWidth = 2
histogramMsrParams = PyIPSDK.createHistoMsrParamsWithBinWidth(binMin, binMax, binWidth)

# computation of reference histogram
refHisto = glbmsr.histogramMsr2d(inImg, histogramMsrParams)
    
# computation of kernel density estimation
kdeDataSet = glbmsr.kernelDensityEstimator2d(inImg)

# extraction of associated histogram
estimHisto = PyIPSDK.generateHistogram(kdeDataSet,
                                       binMin, binMax, binWidth,
                                       nbTotPixels)

We then save the results into output csv files.

# save generated histograms
print("Saving histograms data to directory " + outputDataPath)
PyIPSDK.exportToCsv(refCsvPath, refHisto)
PyIPSDK.exportToCsv(estimCsvPath, estimHisto)

Lastly, if matplotlib python library is available, we plot results into a graph.

    # plot histogram results
    import matplotlib.pyplot as plt
    refPlot, = plt.plot(refHisto.getBinMeanColl(), refHisto.frequencies, 'g+-', label="Reference")
    estimPlot, = plt.plot(estimHisto.getBinMeanColl(), estimHisto.frequencies, 'ro-', label="Estimation")
    plt.title("Image histograms")
    plt.xlabel('Bin mean')
    plt.ylabel('Population')  
    plt.legend(handles=[refPlot, estimPlot])
    plt.grid(True)
    plt.show()

See the full source listing