flowpy_torch.py

This code is a port to PyTorch of the flow to RGB convertion from flowpy.

https://gitlab-research.centralesupelec.fr/2018seznecm/flowpy

ptlflow.utils.flowpy_torch.flow_to_rgb(flow: Tensor, flow_max_radius: float | Tensor | None = None, background: str = 'bright', custom_colorwheel: Tensor | None = None) → Tensor[source]

Create a RGB representation of an optical flow.

Parameters:

flowtorch.Tensor: Flow with at least 3 dimensions in the …CHW (…, Channels, Height, Width) layout, where … represents any number of dimensions. flow[…, 0, h, w] should be the x-displacement flow[…, 1, h, w] should be the y-displacement
flow_max_radiusfloat or torch.Tensor, optional: Set the radius that gives the maximum color intensity, useful for comparing different flows. Default: The normalization is based on the input flow maximum radius per batch element.
backgroundstr, default ‘bright’: States if zero-valued flow should look ‘bright’ or ‘dark’.
custom_colorwheeltorch.Tensor: Use a custom colorwheel for specific hue transition lengths. By default, the default transition lengths are used.

Returns:

torch.Tensor: The RGB representation of the flow. RGB values are float in the [0, 1] interval. The output shape is (…, 3, H, W).

Raises:

ValueError: If the background choice is invalid.