The Detect Image Anomalies tool creates an anomaly score raster from an input multiband or hyperspectral image. The anomaly score raster is a single band raster with values between zero and one.
The output is a hosted imagery layer.
Examples
Detect the number of ships in a satellite image of the ocean, in which the ocean pixels correspond to background values and pixels with ships are anomalies.
Usage notes
Detect Image Anomalies includes configurations for the input layer, anomaly settings, and the result layer.
Input layer
The Input layer group includes the Input raster parameter, which is the multiband or hyperspectral image containing the anomalies to be detected.
Anomaly settings
The Anomaly settings group includes the following parameters:
- Anomaly calculation method determines the method used to detect anomalies. The following options are available:
- RXD (default)—Extracts pixels that are significantly different from the background. This method calculates the pixels' Mahalanobis distance to the background, which is defined by the mean. It has the following formula:
In which r is the sample pixel spectra, µ is the mean spectra, K is covariance, and L is the number of bands.δRXD(r)=(r-µ)TK-1LxL(r-µ) - UTD—Extracts background pixels from the input image. This method is similar to the RXD method but extracts the background. It has the following formula:
Which defines the anomaly by replacing (r-µ) in the RXD method with (1-µ).δUTD(r)=(1-µ)TK-1LxL(1-µ) - KMEANS—Extracts pixels that significantly deviate from established clusters in the data using a K-means clustering algorithm.
- RXD (default)—Extracts pixels that are significantly different from the background. This method calculates the pixels' Mahalanobis distance to the background, which is defined by the mean. It has the following formula:
- Background region specifies the polygon feature layer that defines the region used to calculate background statistics. This parameter only appears if Anomaly calculation method is set to RXD or UTD.
- Number of clusters specifies the number of clusters to use if Anomaly calculation method is set to KMEANS.
- Recompute statistics determines whether to recompute statistics for the input raster when Anomaly calculation method is set to RXD or UTD, which require accurate statistics to perform. If checked, statistics are recomputed for the input raster. This is the default.
Result layer
The Result layer group includes the following parameters:
- Output name specifies the name of the layer that is created and displayed. The name must be unique. If a layer with the same name already exists in your organization, the tool will fail, and you will be prompted to use a different name.
- Output layer type specifies the type of raster output that will be created. The output can be either a tiled imagery layer or a dynamic imagery layer.
- Save in folder specifies the name of a folder in My content where the result will be saved.
Environments
Analysis environment settings are additional parameters that affect a tool's results. You can access the tool's analysis environment settings from the Environment settings parameter group.
This tool honors the following analysis environments:
- Output coordinate system
- Geographic transformations
- Processing extent
Note:
The default processing extent for raster analysis tools and raster functions is Display extent. This default is different from other types of analysis.
- Snap raster
- Cell size
- Resampling method
Credits
This tool consumes credits.
Use Estimate credits to calculate the number of credits that will be required to run the tool. For more information, see Understand credits for spatial analysis.
Outputs
The output is a hosted imagery layer of a single band raster that stores anomaly scores between zero and one as floating point numbers. The background value is zero, and large values approaching one are potential anomaly pixels. You can filter the anomaly pixels using raster functions, such as the Remap or Mask function.
Licensing requirements
This tool requires the following user type and configurations:
- Creator, Professional, or Professional Plus user type
- Publisher, Facilitator, or Administrator role, or an equivalent custom role
Resources
Use the following resources to learn more:
- Detect Image Anomalies in ArcGIS API for Python
- Detect Image Anomalies in ArcGIS Pro