Module ctsimu.evaluation.test2D_FB_2
Test 2D-FB-2: Laws of distance and radiation incidence
This test scenario checks if the intensity profile is correctly rendered under the assumption of an isotropic point source. The inverse square law and the law of radiation incidence must be obeyed:
I\sim \cos(\alpha)/r^2
The test compares the central horizontal row of pixels at index y=250 with the analytical intensity profile.
To run this test evaluation, you can simply pass the name of the test and the metadata file which describes the projection that you want to evaluate to the Toolbox:
from ctsimu.toolbox import Toolbox
Toolbox("2D-FB-2", "2D-FB-2_metadata.json")
Expand source code
# -*- coding: UTF-8 -*-
"""# Test 2D-FB-2: Laws of distance and radiation incidence
.. include:: ./test2D_FB_2.md
"""
from ..test import *
from ..helpers import *
from ..scenario import Scenario
class Test2D_FB_2(generalTest):
""" CTSimU test 2D-FB-2: 1/r^2 law, cos(alpha) intensity on pixel. """
def __init__(self, resultFileDirectory=".", name=None, rawOutput=False):
generalTest.__init__(
self,
testName="2D-FB-2",
name=name,
nExpectedRuns=1,
resultFileDirectory=resultFileDirectory,
rawOutput=rawOutput)
self.geometry = None
self.analyticalIntensityProfileImage = None # stores the image after preparation
# Horizontal profile data:
self.lineNr = 0
self.profile_analytical = None
self.profile_measured = None
# Maximum absolute grey value difference between analytical image and mesured image:
self.maxDiff = 0
# RMS of GV differences:
self.rmsGVDiff = 0
def prepare(self):
""" Preparations before the test will be run with the images from the pipeline. """
if not isinstance(self.pipe, Pipeline):
self.prepared = False
raise Exception("Step must be part of a processing pipeline before it can prepare. Current pipeline: {}".format(self.pipe))
if not self.prepared:
self.jsonScenarioFile = "2D-FB-2_2021-03-24v06r00dp.json"
if(self.jsonScenarioFile is not None):
self.scenario = Scenario(json_dict=json_from_pkg(pkg_scenario(self.jsonScenarioFile)))
self.geometry = self.scenario.current_geometry()
self.geometry.update()
self.analyticalIntensityProfileImage = self.geometry.create_detector_flat_field_analytical()
# Raise normalized image to maximum grey value of 60000,
# as demanded by test 2D-FB-2:
self.analyticalIntensityProfileImage.renormalize(newMin=0, newMax=60000.0, currentMin=0)
self.prepared = True
else:
raise Exception("Test 2D-FB-2: Please provide a JSON scenario description.")
def run(self, image):
self.prepare()
self.currentRun += 1
self.lineNr = int(round(self.geometry.brightest_spot_detector.y()))
log("Getting horizontal profile for detector row {l}.".format(l=self.lineNr))
# Horizontal profile of the analytical image along the central line:
self.profile_analytical = self.analyticalIntensityProfileImage.horizontalProfile(self.lineNr)
# Horizontal profile of the image to be evaluated:
self.profile_measured = image.horizontalProfile(self.lineNr)
return image
def followUp(self):
log("Writing evaluation results...")
nPixels = len(self.profile_analytical)
if len(self.profile_analytical) == len(self.profile_measured):
csvText = "# x [px]\tAnalytical GV\tMeasured GV\tDifference\tRel. Deviation [%]\n"
squaresSum = 0
self.maxDiff = 0
for i in range(nPixels):
gv0 = self.profile_analytical[i]
gv1 = self.profile_measured[i]
delta = gv0 - gv1
relativeDeviation = 0.0
if gv0 != 0:
relativeDeviation = (100.0*delta)/gv0
squaresSum += delta*delta
if abs(delta) > self.maxDiff:
self.maxDiff = abs(delta)
csvText += "{i}\t{gvAnalytical:.3f}\t{gvMeasured:.3f}\t{delta:.3f}\t{relDev:.3f}\n".format(i=i, gvAnalytical=gv0, gvMeasured=gv1, delta=delta, relDev=relativeDeviation)
self.rmsGVDiff = math.sqrt(squaresSum / nPixels)
# Write evaluation text file:
header = "# Evaluation of Test {name}:\n# Max Absolute GV Difference: {maxAbsDiff:.5f}\n# RMS GV Difference: {rms:.5f}\n# \n# Horizontal profile along y={line}:\n".format(name=self.name, maxAbsDiff=self.maxDiff, rms=self.rmsGVDiff, line=self.lineNr)
csvText = header + csvText
csvFileName = "{dir}/{name}_summary.txt".format(dir=self.resultFileDirectory, name=self.name)
with open(csvFileName, 'w') as csvFile:
csvFile.write(csvText)
csvFile.close()
# Write analytical image:
if self.rawOutput:
self.analyticalIntensityProfileImage.saveRAW("{dir}/{name}_analytical.raw".format(dir=self.resultFileDirectory, name=self.name), dataType="float32", addInfo=True)
else: # TIFF
self.analyticalIntensityProfileImage.save("{dir}/{name}_analytical.tif".format(dir=self.resultFileDirectory, name=self.name), dataType="float32")
self.plotResults()
log("Evaluation data for test {name} written to {dir}.".format(name=self.name, dir=self.resultFileDirectory))
else:
raise Exception("Profile width mismatch between analytical result ({}) and measured result ({}).".format(len(profile_analytical), len(profile_measured)))
def plotResults(self):
try:
log("Plotting evaluation results...")
import matplotlib
import matplotlib.pyplot
from matplotlib.ticker import (MultipleLocator, FormatStrFormatter, AutoMinorLocator)
xValues = numpy.linspace(0, len(self.profile_analytical), len(self.profile_analytical), endpoint=False)
absDev = self.profile_analytical - self.profile_measured
relDev = 100.0*absDev / self.profile_analytical
matplotlib.use("agg")
fig, (axUpper, axLower) = matplotlib.pyplot.subplots(nrows=2, ncols=1, figsize=(9, 9))
# Grey Value Profile:
axUpper.plot(xValues, self.profile_analytical, linewidth=8.0, label="Analytical", color='#ffaa00')
axUpper.plot(xValues, self.profile_measured, linewidth=2.0, label="Measured", color='#1f77b4')
axUpper.set_xlabel("Horizontal distance in px")
axUpper.set_ylabel("Grey value")
axUpper.set_title("Grey value line profile along y = {line} px".format(line=self.lineNr))
axUpper.xaxis.set_major_locator(MultipleLocator(100))
axUpper.xaxis.set_major_formatter(FormatStrFormatter('%d'))
axUpper.xaxis.set_minor_locator(MultipleLocator(50))
axUpper.grid(visible=True, which='major', axis='both', color='#d9d9d9', linestyle='dashed')
axUpper.grid(visible=True, which='minor', axis='both', color='#e7e7e7', linestyle='dotted')
axUpper.legend()
# Absolute Deviation:
line_absolute = axLower.plot(xValues, absDev, linewidth=1.25, label="Absolute", color='#ffaa00')
axLower.set_xlabel("Horizontal distance in px")
axLower.set_ylabel("Absolute deviation in grey values")
axLower.set_title("Deviation = Analytical - Measured")
axLower.xaxis.set_major_locator(MultipleLocator(100))
axLower.xaxis.set_major_formatter(FormatStrFormatter('%d'))
axLower.xaxis.set_minor_locator(MultipleLocator(50))
axLower.grid(visible=True, which='major', axis='both', color='#d9d9d9', linestyle='dashed')
axLower.grid(visible=True, which='minor', axis='both', color='#e7e7e7', linestyle='dotted')
# Relative Deviation:
axLower2 = axLower.twinx()
axLower2.axhline(0, linestyle='-', linewidth=1.25, color='#000000')
line_relative = axLower2.plot(xValues, relDev, linewidth=1.25, label="Relative", color='#1f77b4')
axLower2.set_ylabel("Relative deviation in %")
maxDev = max(abs(relDev.min()), abs(relDev.max()))
axLower2.set_ylim([-1.5*maxDev, 1.5*maxDev])
# Add all relevant lines to legend:
lines_all = line_absolute + line_relative
labels = [l.get_label() for l in lines_all]
axLower.legend(lines_all, labels, loc=0)
fig.tight_layout(pad=5.0)
plotFilename = "{dir}/{name}_results.png".format(dir=self.resultFileDirectory, name=self.name)
matplotlib.pyplot.savefig(plotFilename)
fig.clf()
matplotlib.pyplot.close('all')
except Exception as e:
log(f"Warning: Error plotting results for test {self.name}, {subtestName} using matplotlib: {e}")Classes
class Test2D_FB_2 (resultFileDirectory='.', name=None, rawOutput=False)-
CTSimU test 2D-FB-2: 1/r^2 law, cos(alpha) intensity on pixel.
Expand source code
class Test2D_FB_2(generalTest): """ CTSimU test 2D-FB-2: 1/r^2 law, cos(alpha) intensity on pixel. """ def __init__(self, resultFileDirectory=".", name=None, rawOutput=False): generalTest.__init__( self, testName="2D-FB-2", name=name, nExpectedRuns=1, resultFileDirectory=resultFileDirectory, rawOutput=rawOutput) self.geometry = None self.analyticalIntensityProfileImage = None # stores the image after preparation # Horizontal profile data: self.lineNr = 0 self.profile_analytical = None self.profile_measured = None # Maximum absolute grey value difference between analytical image and mesured image: self.maxDiff = 0 # RMS of GV differences: self.rmsGVDiff = 0 def prepare(self): """ Preparations before the test will be run with the images from the pipeline. """ if not isinstance(self.pipe, Pipeline): self.prepared = False raise Exception("Step must be part of a processing pipeline before it can prepare. Current pipeline: {}".format(self.pipe)) if not self.prepared: self.jsonScenarioFile = "2D-FB-2_2021-03-24v06r00dp.json" if(self.jsonScenarioFile is not None): self.scenario = Scenario(json_dict=json_from_pkg(pkg_scenario(self.jsonScenarioFile))) self.geometry = self.scenario.current_geometry() self.geometry.update() self.analyticalIntensityProfileImage = self.geometry.create_detector_flat_field_analytical() # Raise normalized image to maximum grey value of 60000, # as demanded by test 2D-FB-2: self.analyticalIntensityProfileImage.renormalize(newMin=0, newMax=60000.0, currentMin=0) self.prepared = True else: raise Exception("Test 2D-FB-2: Please provide a JSON scenario description.") def run(self, image): self.prepare() self.currentRun += 1 self.lineNr = int(round(self.geometry.brightest_spot_detector.y())) log("Getting horizontal profile for detector row {l}.".format(l=self.lineNr)) # Horizontal profile of the analytical image along the central line: self.profile_analytical = self.analyticalIntensityProfileImage.horizontalProfile(self.lineNr) # Horizontal profile of the image to be evaluated: self.profile_measured = image.horizontalProfile(self.lineNr) return image def followUp(self): log("Writing evaluation results...") nPixels = len(self.profile_analytical) if len(self.profile_analytical) == len(self.profile_measured): csvText = "# x [px]\tAnalytical GV\tMeasured GV\tDifference\tRel. Deviation [%]\n" squaresSum = 0 self.maxDiff = 0 for i in range(nPixels): gv0 = self.profile_analytical[i] gv1 = self.profile_measured[i] delta = gv0 - gv1 relativeDeviation = 0.0 if gv0 != 0: relativeDeviation = (100.0*delta)/gv0 squaresSum += delta*delta if abs(delta) > self.maxDiff: self.maxDiff = abs(delta) csvText += "{i}\t{gvAnalytical:.3f}\t{gvMeasured:.3f}\t{delta:.3f}\t{relDev:.3f}\n".format(i=i, gvAnalytical=gv0, gvMeasured=gv1, delta=delta, relDev=relativeDeviation) self.rmsGVDiff = math.sqrt(squaresSum / nPixels) # Write evaluation text file: header = "# Evaluation of Test {name}:\n# Max Absolute GV Difference: {maxAbsDiff:.5f}\n# RMS GV Difference: {rms:.5f}\n# \n# Horizontal profile along y={line}:\n".format(name=self.name, maxAbsDiff=self.maxDiff, rms=self.rmsGVDiff, line=self.lineNr) csvText = header + csvText csvFileName = "{dir}/{name}_summary.txt".format(dir=self.resultFileDirectory, name=self.name) with open(csvFileName, 'w') as csvFile: csvFile.write(csvText) csvFile.close() # Write analytical image: if self.rawOutput: self.analyticalIntensityProfileImage.saveRAW("{dir}/{name}_analytical.raw".format(dir=self.resultFileDirectory, name=self.name), dataType="float32", addInfo=True) else: # TIFF self.analyticalIntensityProfileImage.save("{dir}/{name}_analytical.tif".format(dir=self.resultFileDirectory, name=self.name), dataType="float32") self.plotResults() log("Evaluation data for test {name} written to {dir}.".format(name=self.name, dir=self.resultFileDirectory)) else: raise Exception("Profile width mismatch between analytical result ({}) and measured result ({}).".format(len(profile_analytical), len(profile_measured))) def plotResults(self): try: log("Plotting evaluation results...") import matplotlib import matplotlib.pyplot from matplotlib.ticker import (MultipleLocator, FormatStrFormatter, AutoMinorLocator) xValues = numpy.linspace(0, len(self.profile_analytical), len(self.profile_analytical), endpoint=False) absDev = self.profile_analytical - self.profile_measured relDev = 100.0*absDev / self.profile_analytical matplotlib.use("agg") fig, (axUpper, axLower) = matplotlib.pyplot.subplots(nrows=2, ncols=1, figsize=(9, 9)) # Grey Value Profile: axUpper.plot(xValues, self.profile_analytical, linewidth=8.0, label="Analytical", color='#ffaa00') axUpper.plot(xValues, self.profile_measured, linewidth=2.0, label="Measured", color='#1f77b4') axUpper.set_xlabel("Horizontal distance in px") axUpper.set_ylabel("Grey value") axUpper.set_title("Grey value line profile along y = {line} px".format(line=self.lineNr)) axUpper.xaxis.set_major_locator(MultipleLocator(100)) axUpper.xaxis.set_major_formatter(FormatStrFormatter('%d')) axUpper.xaxis.set_minor_locator(MultipleLocator(50)) axUpper.grid(visible=True, which='major', axis='both', color='#d9d9d9', linestyle='dashed') axUpper.grid(visible=True, which='minor', axis='both', color='#e7e7e7', linestyle='dotted') axUpper.legend() # Absolute Deviation: line_absolute = axLower.plot(xValues, absDev, linewidth=1.25, label="Absolute", color='#ffaa00') axLower.set_xlabel("Horizontal distance in px") axLower.set_ylabel("Absolute deviation in grey values") axLower.set_title("Deviation = Analytical - Measured") axLower.xaxis.set_major_locator(MultipleLocator(100)) axLower.xaxis.set_major_formatter(FormatStrFormatter('%d')) axLower.xaxis.set_minor_locator(MultipleLocator(50)) axLower.grid(visible=True, which='major', axis='both', color='#d9d9d9', linestyle='dashed') axLower.grid(visible=True, which='minor', axis='both', color='#e7e7e7', linestyle='dotted') # Relative Deviation: axLower2 = axLower.twinx() axLower2.axhline(0, linestyle='-', linewidth=1.25, color='#000000') line_relative = axLower2.plot(xValues, relDev, linewidth=1.25, label="Relative", color='#1f77b4') axLower2.set_ylabel("Relative deviation in %") maxDev = max(abs(relDev.min()), abs(relDev.max())) axLower2.set_ylim([-1.5*maxDev, 1.5*maxDev]) # Add all relevant lines to legend: lines_all = line_absolute + line_relative labels = [l.get_label() for l in lines_all] axLower.legend(lines_all, labels, loc=0) fig.tight_layout(pad=5.0) plotFilename = "{dir}/{name}_results.png".format(dir=self.resultFileDirectory, name=self.name) matplotlib.pyplot.savefig(plotFilename) fig.clf() matplotlib.pyplot.close('all') except Exception as e: log(f"Warning: Error plotting results for test {self.name}, {subtestName} using matplotlib: {e}")Ancestors
Methods
def prepare(self)-
Preparations before the test will be run with the images from the pipeline.
Expand source code
def prepare(self): """ Preparations before the test will be run with the images from the pipeline. """ if not isinstance(self.pipe, Pipeline): self.prepared = False raise Exception("Step must be part of a processing pipeline before it can prepare. Current pipeline: {}".format(self.pipe)) if not self.prepared: self.jsonScenarioFile = "2D-FB-2_2021-03-24v06r00dp.json" if(self.jsonScenarioFile is not None): self.scenario = Scenario(json_dict=json_from_pkg(pkg_scenario(self.jsonScenarioFile))) self.geometry = self.scenario.current_geometry() self.geometry.update() self.analyticalIntensityProfileImage = self.geometry.create_detector_flat_field_analytical() # Raise normalized image to maximum grey value of 60000, # as demanded by test 2D-FB-2: self.analyticalIntensityProfileImage.renormalize(newMin=0, newMax=60000.0, currentMin=0) self.prepared = True else: raise Exception("Test 2D-FB-2: Please provide a JSON scenario description.")
Inherited members