swectral.functions.aucnorm_hyper

swectral.functions.aucnorm_hyper(data)

AUC (Area Under Curve) normalization function for image pixel spectrum correction in SpecPipe pipelines. This function is for optimized GPU-accelerated application. :rtype: Tensor

Process input data level: 4 - ‘pixel_hyperspecs_tensor’

Process output data level: 4 - ‘pixel_hyperspecs_tensor’

Parameters:

data3D array_like (n_samples, n_bands, 1)

Three-dimensional array containing 2D spectral data to be processed. The size of the third dimension must be exactly 1.

Returns:

torch.Tensor

AUC normalization transformed spectral data.

Examples

>>> aucnorm_hyper([[[1], [2], [3], [4], [5], [6]], [[2], [2], [4], [4], [6], [6]]])

Incorporation into pipeline for image processing, for SpecPipe instance pipe:

>>> pipe.add_process(4, 4, 0, aucnorm_hyper)