The Detect Objects Using Deep Learning tool uses a deep learning model to identify and locate objects in an imagery layer.
The output is a hosted feature layer.
Examples
Example scenarios for the use of this tool include the following:
- Identify building footprints to upgrade property tax data for a local government or regional emergency response group. The output layer from the tool is a feature layer that can identify the buildings in an area. The feature layer created can be used to match existing property records to record the current building footprint of the property.
- Identify cars in a parking lot to count attendance and prepare traffic surveys. The feature layer created can be used in the Classify Objects Using Deep Learning tool to classify the type of car detected.
Usage notes
Detect Objects Using Deep Learning includes configurations for the input layer, model settings, and the result layer.
Input layer
The Input layer group includes the following parameters:
- Input imagery layer or feature layer is the imagery layer or feature layer with attachments that will be used to detect the objects identified in the deep learning model. The imagery layer selected should be based on the requirements of the deep learning model that will be used to classify the pixels.
- Processing mode specifies how the raster items in the imagery layer will be processed. The options are as follows:
- Process as mosaicked image—All raster items in the mosaic dataset or image service will be mosaicked together and processed. This is the default.
- Process all raster items separately—All raster items in the mosaic dataset or image service will be processed as separate images.
Model settings
The Model settings group includes the following parameters:
- Model for object detection is the deep learning model that will be used to detect the objects. The deep learning model must be located on ArcGIS Online to be selected in the tool. You can select your own model, a publicly available model in ArcGIS Online, or a model from ArcGIS Living Atlas of the World.
- Model arguments specifies the function arguments defined in the Python raster function class. Additional deep learning parameters and arguments for experiments and refinement are listed, such as a confidence threshold for adjusting the sensitivity. The names of the arguments are populated from the Python module.
- Non maximum suppression (NMS) specifies whether nonmaximum suppression will be performed to remove duplicate objects that are identified based on confidence values.
- Confidence score field is the field name that will record the confidence scores that are created as output by the object detection method. This parameter is available when Non maximum suppression (NMS) is enabled.
- Class value field is the field in the output feature layer that will contain the value from the input imagery layer. If no value is specified, the standard class value fields Classvalue and Value will be used. If these fields do not exist, all features will be treated as the same object class. This parameter is available when Non maximum suppression (NMS) is enabled.
- Maximum overlap ratio specifies the ratio of intersection area over the union area for two overlapping features. The default value is 0. This parameter is available when Non maximum suppression (NMS) is enabled.
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.
- 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 is Full extent. This default is different from Map Viewer Classic in which Use current map extent is enabled by default.
- Cell size
- Mask
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 feature layer with each detected object as an individual feature with the class value and confidence fields added.
Usage requirements
This tool requires the following user type and configurations:
- Professional or Professional Plus user type
- Publisher, Facilitator, or Administrator role, or an equivalent custom role with the Imagery Analysis privilege
Resources
Use the following resources to learn more:
- Detect Objects Using Deep Learning in ArcGIS REST API
- detect_objects function in ArcGIS API for Python
- Classify Objects Using Deep Learning in ArcGIS Online
- Detect Objects Using Deep Learning in ArcGIS Pro