dvpio.read.shapes.read_lmd

dvpio.read.shapes.read_lmd(path, calibration_points_image, transformation_type='similarity', precision=6, switch_orientation=False)

Read and parse LMD-formatted masks for the use in spatialdata

Wrapper for pyLMD functions.

Parameters:

• path (str) – Path to LMD-formatted segmentation masks in .xml format

• calibration_points_image (PointsModel) – Calibration points of the image as DataFrame, with 3 calibration points. Point coordinates are stored as seperate columns in x and y column.

• transformation_type (Literal['similarity', 'affine', 'euclidean'] (default: 'similarity')) –

  • affine

    Full affine transformation (scaling, rotation/reflexion, translation, shearing). This operation does not preserve the angles within or distances the shapes

  • similarity (recommended)

    Similarity transformation. Compared to an affine transformation, a similarity transformation constraints the solution space to scaling, rotations, reflections, and translations, i.e. angles of shapes are retained. If you only want to map between image and microscopy coordinates only the subset of similarity transformations (scaling, rotation, reflection, translation) is required.

  • euclidean (Rigid transform)

    Only translation and rotation are allowed

• precision (int | None (default: 6)) – Default 6. Rounding of affine transformation matrix, which can be necessary for numerical stability of shape transformations. Passing None skips rounding.

• switch_orientation (bool (default: False)) – Per default, LMD is working in a (x, y) coordinate system while the image coordinates are in a (row=y, col=x) coordinate system. If True, transform the coordinate systems by mirroring the coordinate system at the main diagonal.

Return type:

ShapesModel

Returns:

spatialdata.models.ShapesModel Transformed shapes in image coordinates.

Object has special attributes

• attrs.transformation

  • global (image coordinates)

  • to_lmd Transformation back to leica coordinate system