Type Members

1.  case class ASMSample(mesh: TriangleMesh[_3D], featureField: DiscreteFeatureField[_3D, UnstructuredPointsDomain], featureExtractor: FeatureExtractor) extends Product with Serializable

   Class of instances sampled from an Active Shape Model.

   Class of instances sampled from an Active Shape Model. A sample is therefore comprised of both a shape sampled from the Shape Model, and a set of sample features at the profile points.

2.  case class ActiveShapeModel(statisticalModel: StatisticalMeshModel, profiles: Profiles, preprocessor: ImagePreprocessor, featureExtractor: FeatureExtractor) extends Product with Serializable
3.  class DiscreteFeatureField[D, DDomain[D] <: DiscreteDomain[D]] extends DiscreteField[D, DDomain, DenseVector[Double]]

   Class of Discrete Fields where to each point, a DenseVector of arbitrary dimensionality is associated.

   Class of Discrete Fields where to each point, a DenseVector of arbitrary dimensionality is associated.

   An example instance of such a class is the set of profile points and associated features in an Active Shape Model.

4.  trait FeatureExtractor extends (PreprocessedImage, Point[_3D], TriangleMesh[_3D], PointId) => Option[DenseVector[Double]] with HasIOMetadata
5.  trait FeatureExtractorIOHandler extends IOHandler[FeatureExtractor]
6.  case class FittingConfiguration(featureDistanceThreshold: Double, pointDistanceThreshold: Double, modelCoefficientBounds: Double) extends Product with Serializable

   Fitting Configuration, specifying thresholds and bounds.

   Fitting Configuration, specifying thresholds and bounds.

   featureDistanceThreshold

   threshold for the feature distance. If the mahalanobis distance of a candidate point's features to the corresponding profile's mean is larger than this value, then that candidate point will be ignored during fitting.

   pointDistanceThreshold

   threshold for point distance: If the mahalanobis distance of a candidate point to its corresponding marginal distribution is larger than this value, then that candidate point will be ignored during fitting.

   modelCoefficientBounds

   bounds to apply on the model coefficients. In other words, by setting this to n, all coefficients of the fitting result will be restricted to the interval [-n, n].

7.  case class FittingResult(transformations: ModelTransformations, mesh: TriangleMesh[_3D]) extends Product with Serializable

   Fitting results.

   Fitting results.

   Note that the fields are redundant: the mesh is completely determined by the transformations. It's essentially provided for user convenience, because it would be very likely to be (re-)constructed anyway from the transformations.

   transformations

   transformations to apply to the model

   mesh

   the mesh resulting from applying these transformations

8.  case class GaussianGradientImagePreprocessor(stddev: Double, ioMetadata: IOMetadata = GaussianGradientImagePreprocessor.IOMetadata_Default) extends ImagePreprocessor with Product with Serializable

   Image preprocessor that calculates a gradient image from the input image.

   Image preprocessor that calculates a gradient image from the input image. The following steps are performed:

   1. The image is filtered using a Gaussian filter. The sigma parameter used for the filter is passed through from stddev. If stddev is 0, then no blurring is performed. 2. The resulting image is interpolated (using B-Spline interpolation of order 1). 3. The resulting image is differentiated to produce a gradient image.

   stddev

   the standard deviation (in millimeters) to use for the gaussian blur filter. Set to 0 to disable blurring.

   ioMetadata

   IO Metadata

9.  trait HasIOIdentifier extends AnyRef

   A trait signifying that the implementing object provides a unique identifier used for serialization purposes.

10.  trait HasIOMetadata extends AnyRef

    A trait signifying that the implementing object provides identity and version information for (de)serialization purposes.

11.  trait Hdf5IOHandler[T <: HasIOMetadata] extends AnyRef

    Trait providing methods for serializing/deserializing objects of type T to/from HDF5 files.

    Trait providing methods for serializing/deserializing objects of type T to/from HDF5 files.

    T

    the type of objects which can be constructed from the information present in HDF5 files, and saved into such files.

12.  trait IOHandler[T <: HasIOMetadata] extends Hdf5IOHandler[T] with HasIOIdentifier

    Trait specifying that the implementing object can load and save objects of a given type from/to files.

    Trait specifying that the implementing object can load and save objects of a given type from/to files. In addition, the object specifies which IO Identifier it can handle.

    Currently, we require that the HDF5 file format be supported.

    T

    the type of objects which can be constructed from the information present in files, and saved into files.

    See also

    IOCollection

13.  class IOHandlers[T <: HasIOMetadata, IO <: IOHandler[T]] extends AnyRef

    An object representing a collection of IO handlers providing load/save capabilities for a given object type.

    An object representing a collection of IO handlers providing load/save capabilities for a given object type. If there are multiple implementations for the given type (e.g., multiple image preprocessor implementations, multiple feature extractor implementations), they are distinguished (and uniquely identified) by their IO Identifiers (see HasIOIdentifier).

    For every supported IO Identifier, a corresponding IOHandler must be registered using the register() method. The identifiers/handlers that come built into scalismo are automatically registered and available, but manual registration is required for user-defined handlers.

    T

    the type of objects that can be loaded/saved

    IO

    the type of the corresponding IO handlers

14.  case class IOMetadata(identifier: String, majorVersion: Int, minorVersion: Int) extends HasIOIdentifier with Product with Serializable

    Metadata about an object, used for serialization purposes.

    Metadata about an object, used for serialization purposes. In addition to the unique identifier, attributes concerning the implementation version are provided.

    identifier

    a unique IO identifier.

    majorVersion

    major implementation version.

    minorVersion

    minor implementation version.

    See also

    HasIOIdentifier

15.  case class IdentityImagePreprocessor(ioMetadata: IOMetadata = IdentityImagePreprocessor.IOMetadata_Default) extends ImagePreprocessor with Product with Serializable

    The "identity" Preprocessor performs no preprocessing.

    The "identity" Preprocessor performs no preprocessing. In other words, this class can be considered an adapter that turns a discrete image into a PreprocessedImage, without modifying the image.

    ioMetadata

    IO Metadata

16.  trait ImagePreprocessor extends (DiscreteImage[_3D, Float]) => PreprocessedImage with HasIOMetadata

    An image preprocessor takes a discrete scalar image, performs any required preprocessing, and returns a PreprocessedImage which will serve as input to a FeatureExtractor.

    An image preprocessor takes a discrete scalar image, performs any required preprocessing, and returns a PreprocessedImage which will serve as input to a FeatureExtractor.

    When implementing a custom preprocessor, make sure to define and register an accompanying IO Handler.

    See also

    ImagePreprocessorIOHandler

    ImagePreprocessorIOHandlers

17.  trait ImagePreprocessorIOHandler extends IOHandler[ImagePreprocessor]

    IO Handler for the ImagePreprocessor type.

18.  case class ModelTransformations(coefficients: DenseVector[Double], rigidTransform: TranslationAfterRotation[_3D]) extends Product with Serializable

    Transformations to apply to a statistical shape model.

    Transformations to apply to a statistical shape model.

    Sample usage: val mesh = ssm.instance(t.coefficients).transform(t.rigidTransform)

    coefficients

    model coefficients to apply. These determine the shape transformation.

    rigidTransform

    rigid transformation to apply. These determine translation and rotation.

19.  case class NormalDirectionFeatureExtractor(numberOfPoints: Int, spacing: Double, ioMetadata: IOMetadata = NormalDirectionFeatureExtractor.IOMetadata_Default) extends FeatureExtractor with Product with Serializable
20.  case class NormalDirectionSearchPointSampler(numberOfPoints: Int, searchDistance: Float) extends SearchPointSampler with Product with Serializable
21.  trait PreprocessedImage extends Field[_3D, DenseVector[Float]]
22.  case class Profile(pointId: PointId, distribution: MultivariateNormalDistribution) extends Product with Serializable
23.  final case class ProfileId(id: Int) extends AnyVal with Product with Serializable
24.  case class Profiles(data: IndexedSeq[Profile]) extends Iterable[Profile] with Product with Serializable
25.  trait SearchPointSampler extends (TriangleMesh[_3D], PointId) => Seq[Point[_3D]]