similarity_2d_instance_model¶

menpofit.modelinstance.similarity_2d_instance_model(shape)[source]¶

Creates a menpo.model.MeanLinearModel that encodes the 2D similarity transforms that can be applied on a 2D shape that consists of n_points.

Parameters: shape (menpo.shape.PointCloud) – The input 2D shape.
Returns: model (subclass of menpo.model.MeanLinearModel) – Linear model with four components, the linear combinations of which represent the original shape under a similarity transform. The model is exhaustive (that is, all possible similarity transforms can be expressed with the model).