class Spectral_data

A PlantCV data object class.

class plantcv.Spectral_data

Spectral_data is a class that is used to create instances of hyperspectral/multispectral data objects for PlantCV analysis. An instance of Spectral_data is created automatically as hyperspectral data is read in with pcv.readimage. Functions from the PlantCV hyperspectral sub-package utilize the plantcv.Spectral_data instance attributes, which has metadata about the hyperspectral data instance, to simplify function input. Most of this metadata is collected from the .hdr file. These attributes are used internally by PlantCV functions but also can be utilized by users.

Attributes

Attributes are accessed as spectral_data_instance.attribute.

array_data: The actual data, stored as a Numpy array.

max_wavelength: Largest available wavelength in the spectral datacube.

min_wavelength: Smallest available wavelength in the spectral datacube.

d_type: Data type of the Numpy array data

wavelength_dict: The wavelength dictionary that gets created during pcv.readimage while mode='envi' since metadata is collected from the .hdr file

samples: Width of a single band of spectral data, shaped (samples, lines) rather than (x,y) with other arrays.

lines: Length of a single band of spectral data, shaped (samples, lines) rather than (x,y) with other arrays.

interleave: Interleave type

wavelength_units: Units of the wavelengths

array_type: The type of array data (entire datacube, specific index, first derivative, etc)

filename: The filename where the data originated from

default_bands: The bands that are used to make the pseudo-RGB image, read from the default bands field of the .hdr file. Contains one band number for a grayscale image or three band numbers for an RGB image. Set to None if the .hdr file does not name any default bands, in which case the bands are selected from the available wavelengths.

byte_order: Order of bytes per ENVI standards, defaults to "Network". This is not used by PlantCV but is stored for cross platform compatibility.

file_type: Type of raster per ENVI standards, defaults to "ENVI". This is not used by PlantCV but is stored for cross platform compatibility.

header_offset: Number of bytes ENVI should skip when reading the file, defaults to 0. This is not used by PlantCV but is stored for cross platform compatibility.

metadata: Metadata in a dictionary. Included keys might be specific to hyperspectral or geospatial image types

Methods

Methods are accessed as object_name.method.

pseudo_rgb(): Calculates and returns a pseudo-RGB image. If fewer than 3 wavelengths are present then the first array is repeated to make a 3 channel image.

Example

PlantCV functions from the hyperspectral sub-package use Spectral_data implicitly.

from plantcv import plantcv as pcv


# Read in data
spectral_data_instance = pcv.readimage(filename="ee362e84-c861-4f0d-abbb-215ac58eed60_data", mode='envi')

index_array_gdvi = pcv.hyperspectral.extract_index(array=spectral_data_instance, 
                                                   index="GDVI",
                                                   distance=20)

# Plot the pseudo-RGB image that is created when reading in hyperspectral data 
pcv.plot_image(spectral_data_instance.pseudo_rgb())

# Plot the index extracted from the hyperspectral datacube 
pcv.plot_image(index_array_gdvi.array_data)

print(str(spectral_data_instance.max_wavelength) + spectral_data_instance.wavelength_units)

Psuedo-RGB image from hyperspectral datacube

Screenshot

GDVI Grayscale Image

Screenshot

Maximum wavelength

> 1000.95nm

Source Code: Here