static void OnOptimizerIterationEvent(itkObject sender, itkEventArgs e)
{
OptimizerType optimizer = sender as OptimizerType;
String message = "{0}: {1}";
String iteration = optimizer.CurrentIteration.ToString("000");
String position = optimizer.CurrentPosition.ToString();
Console.WriteLine(String.Format(message, iteration, position));
}
static void Main(string[] args) {
try {
// Get input parameters
if (args.Length < 3) {
Console.WriteLine("Missing Parameters");
Console.Write("Usage: " + Environment.GetCommandLineArgs()[0] + " ");
Console.Write("fixedImageFile movingImageFile ");
Console.Write("outputImagefile ");
Console.WriteLine("[checkerBoardBefore] [checkerBoardAfter]");
return;
}
// Create components
TransformType transform = TransformType.New();
OptimizerType optimizer = OptimizerType.New();
InterpolatorType interpolator = InterpolatorType.New();
MetricType metric = MetricType.New();
RegistrationType registration = RegistrationType.New();
registration.SetOptimizer(optimizer);
registration.SetTransform(transform);
registration.SetInterpolator(interpolator);
registration.SetMetric(metric);
// Set metric parameters
metric.FixedImageStandardDeviation = 0.4;
metric.MovingImageStandardDeviation = 0.4;
// Read input images
ImageType fixedImage = ImageType.New();
fixedImage.Read(args[0]);
ImageType movingImage = ImageType.New();
movingImage.Read(args[1]);
// Normailize
NormalizeFilterType fixedNormalizer = NormalizeFilterType.New();
NormalizeFilterType movingNormalizer = NormalizeFilterType.New();
GaussianFilterType fixedSmoother = GaussianFilterType.New();
GaussianFilterType movingSmoother = GaussianFilterType.New();
fixedSmoother.Variance = new double[] { 2.0, 2.0 };
movingSmoother.Variance = new double[] { 2.0, 2.0 };
// Setup pipeline
fixedNormalizer.SetInput(fixedImage);
movingNormalizer.SetInput(movingImage);
fixedSmoother.SetInput(fixedNormalizer.GetOutput());
movingSmoother.SetInput(movingNormalizer.GetOutput());
ImageType fixedSmootherOutput = ImageType.New();
fixedSmoother.GetOutput(fixedSmootherOutput);
ImageType movingSmootherOutput = ImageType.New();
movingSmoother.GetOutput(movingSmootherOutput);
registration.SetFixedImage(fixedSmootherOutput);
registration.SetMovingImage(movingSmootherOutput);
// Set requested region
fixedNormalizer.Update();
ImageType fixedNormalizerOutput = ImageType.New();
fixedNormalizer.GetOutput(fixedNormalizerOutput);
RegionType fixedImageRegion = fixedNormalizerOutput.BufferedRegion;
fixedImage.RequestedRegion = fixedImageRegion;
//registration.FixedImageRegion = fixedImageRegion
// Set initial parameters
ParametersType initialParameters =
new ParametersType(transform.NumberOfParameters);
initialParameters[0] = 0.0;
initialParameters[1] = 0.0;
registration.InitialTransformParameters = initialParameters;
// Set metric sampling
int numberOfPixels = fixedImageRegion.Size[0] * fixedImageRegion.Size[1];
int numberOfSamples = (int)((double)numberOfPixels * 0.01);
metric.NumberOfSpatialSamples = (uint)numberOfSamples;
// Setup optimizer
optimizer.LearningRate = 15.0;
optimizer.NumberOfIterations = 200;
optimizer.Maximize = true;
// Listen for iteration events
optimizer.Iteration += OnOptimizerIterationEvent;
// Start registration
registration.StartRegistration();
Console.WriteLine(optimizer.StopCondition);
// Get results
ParametersType finalParameters = registration.LastTransformParameters;
double TranslationAlongX = finalParameters[0];
double TranslationAlongY = finalParameters[1];
uint numberOfIterations = optimizer.NumberOfIterations;
double bestValue = optimizer.GetValue();
// Print results
Console.WriteLine("Result=");
Console.WriteLine(String.Format(" Translation X = {0}", TranslationAlongX));
Console.WriteLine(String.Format(" Translation Y = {0}", TranslationAlongY));
Console.WriteLine(String.Format(" Iterations = {0}", numberOfIterations));
Console.WriteLine(String.Format(" Metric value = {0}", bestValue));
Console.WriteLine(String.Format(" Numb. Samples = {0}", numberOfSamples));
// Setup resample filter
TransformType finalTransform = TransformType.New();
finalTransform.Parameters = finalParameters;
//finalTransform.FixedParameters = transform.FixedParameters;
ResampleFilterType resample = ResampleFilterType.New();
resample.SetTransform(finalTransform);
resample.SetInput(movingImage);
resample.OutputSize = fixedImage.LargestPossibleRegion.Size;
resample.OutputSpacing = fixedImage.Spacing;
resample.OutputOrigin = fixedImage.Origin;
resample.DefaultPixelValue = 100;
// Write resampled output
OutputImageType output = OutputImageType.New();
CastFilterType caster = CastFilterType.New(movingImage, output);
caster.SetInput(resample.GetOutput());
caster.Update();
caster.GetOutput(output);
output.Write(args[2]);
// Generate checkerboards before registration
CheckerBoardType checker = CheckerBoardType.New();
checker.SetInput1(fixedImage);
checker.SetInput2(movingImage);
checker.Update();
ImageType checkerOut = ImageType.New();
checker.GetOutput(checkerOut);
caster.SetInput(checkerOut);
caster.Update();
caster.GetOutput(output);
if (args.Length > 3) {
output.Write(args[3]);
}
// Generate checkerboards after registration
checker.SetInput1(fixedImage);
checker.SetInput2(resample.GetOutput());
checker.Update();
checker.GetOutput(checkerOut);
caster.SetInput(checkerOut);
caster.Update();
caster.GetOutput(output);
if (args.Length > 4) {
output.Write(args[4]);
}
}
catch (Exception ex)
{
Console.WriteLine(ex.ToString());
}
}
