Linear Spectral Unmixing (Image Analyst)

Available with Spatial Analyst license.

Summary

Performs subpixel classification and calculates the fractional abundance of different land-cover types for individual pixels.

Usage

• This tool calculates the fractional cover for individual pixels that contain multiple land-cover types. It generates a multiband raster in which each band corresponds to the fractional abundance of each land-cover class. For example, you can use it to perform land-cover classification on a multispectral image to identify photosynthetic vegetation, bare soil, and dead or nonphotosynthetic vegetation.

• The order of the output multiband raster follows the order of the input spectral profile.

• The number of classes cannot exceed the number of bands in the input raster. For example, you cannot extract information about more than eight classes from an 8-band raster.

• The following is an example of a spectral profile provided as a .json file:

  
  {
    "EsriEndmemberDefinitionFile" : 0,
    "FileVersion" : 1,
    "NumberOfEndmembers" : 3,
    "NumberOfBands" : 7,
    "Endmembers" : [	
      {
        "EndmemberID" : 1,
        "EndmemberName" : "urban",
        "SpectralProfile" : [
              88,
  			                     42,
  			                     48,
  			                     38,
  			                     86,
  			                    115,
  			                     59
            ]
      },
      {
        "EndmemberID" : 2,
        "EndmemberName" : "vegetation",
        "SpectralProfile" : [
  			                       50,
  			                       21,
  			                       20,
  			                       35,
  			                       50,
  			                      110,
  			                       23
            ]
      },
      {
        "EndmemberID" : 3,
        "EndmemberName" : "water",
        "SpectralProfile" : [
  			                       51,
  			                       20,
  			                       14,
  			                        9,
  			                        7,
  			                      116,
  			                        4
            ]
      }
    ]        
  }

• Classifier definition files (.ecd) generated from the Train Maximum Likelihood Classifier tool are the only classifier output currently supported.

• Polygon features require the following field names:

  • classname—A text field indicating the name of the class category
  • classvalue—A long integer field containing the integer value for each class category

• When calculating the fractional abundance of each land-cover class, the solution can include negative coefficients or fractions. If this occurs, review the training samples in the input spectral profile to confirm that they accurately represent each class. If they appear correct, select Non-negative for Output Value Option.

Parameters

Dialog

Label	Explanation	Data Type
Input Raster	The input raster dataset.	Raster Dataset; Mosaic Dataset; Mosaic Layer; Raster Layer; File; Image Service
Input Training Features or Spectral Profile	The spectral information for the different land-cover classes. This can be provided as polygon features, a training sample feature class generated from the Training Samples Manager pane, a classifier definition file (.ecd) generated from the Train Maximum Likelihood Classifier tool, or a JSON format file (.json) that contains the class spectral profiles.	File; Feature Layer; String
Output Value Option (Optional)	Specifies how the output pixel values will be defined. Sum to one—Class values for each pixel will be provided in decimal format with the sum of all classes equal to 1. For example, Class1 = 0.16; Class2 = 0.24; Class3 = 0.60. Non-negative—There will be no negative output values.	String

Return Value

Label	Explanation	Data Type
Output Raster	The output multiband raster dataset.	Raster

Python

LinearSpectralUnmixing(in_raster, in_spectral_profile_file, {value_option})

Name	Explanation	Data Type
in_raster	The input raster dataset.	Raster Dataset; Mosaic Dataset; Mosaic Layer; Raster Layer; File; Image Service
in_spectral_profile_file	The spectral information for the different land-cover classes. This can be provided as polygon features, a classifier definition file (.ecd) generated from the Train Maximum Likelihood Classifier tool, or a JSON format file (.json) that contains the class spectral profiles.	File; Feature Layer; String
value_option [value_option,...] (Optional)	Specifies how the output pixel values will be defined. SUM_TO_ONE—Class values for each pixel will be provided in decimal format with the sum of all classes equal to 1. For example, Class1 = 0.16; Class2 = 0.24; Class3 = 0.60. NON_NEGATIVE—There will be no negative output values.	String

Return Value

Name	Explanation	Data Type
out_raster	The output multiband raster dataset.	Raster

Code sample

LinearSpectralUnmixing example 1 (Python window)

This example calculates the fractional abundance of classes from a classifier definition file (.ecd) and generates a multiband raster.

# Import system modules
import arcpy
from arcpy.ia import *

# Check out the ArcGIS Image Analyst extension license
arcpy.CheckOutExtension("ImageAnalyst")

# Execute 
unmixing_outputs = LinearSpectralUnmixing("C:/data/landsat7_image.crf",
    "C:/data/train_maxi_likelihood_ecd_output.ecd", "SUM_TO_ONE;NON_NEGATIVE")
	
# Save output
unmixing_outputs.save("C:/data/unmixing_outputs.crf")

LinearSpectralUnmixing example 2 (stand-alone script)

This example calculates the fractional abundance of classes from a classifier definition file (.ecd) and generates a multiband raster.

# Import system modules
import arcpy
from arcpy.ia import *

# Check out the ArcGIS Image Analyst extension license
arcpy.CheckOutExtension("ImageAnalyst")

# Define input parameters
inFile = "C:/data/landsat7_image.crf"
json_file = "C:/data/customized_endmembers.json"
options = "SUM_TO_ONE" 

# Execute 
unmixing_outputs = LinearSpectralUnmixing(inFile, json_file, options)
	
# Save output
unmixing_outputs.save("C:/data/unmixing_outputs_using_json.crf")

LinearSpectralUnmixing example 3 (stand-alone script)

This example calculates the fractional abundance of classes from a training sample feature class (.shp) and generates a multiband raster.

# Import system modules
import arcpy
from arcpy.ia import *

# Check out the ArcGIS Image Analyst extension license
arcpy.CheckOutExtension("ImageAnalyst")

# Define input parameters
inFile = "C:/data/landsat7_image.crf"
training_features = "C:/data/training_features.shp"
options = "SUM_TO_ONE;NON_NEGATIVE" 

# Execute 
unmixing_outputs = LinearSpectralUnmixing(inFile, training_features, options)
	
# Save output
unmixing_outputs.save("C:/data/unmixing_outputs_using_training_features.crf")

Environments

Compression, Current Workspace, Geographic Transformations, NoData, Output CONFIG Keyword, Output Coordinate System, Extent, Parallel Processing Factor, Pyramid, Raster Statistics, Resampling Method, Scratch Workspace, Snap Raster, Tile Size