AbstractImageAnalyzer QML Type

(BETA) Provides base functionality to perform image classification or object detection. More...

Import Statement: import ArcGIS.AppFramework.DeepLearning 1.0
Inherited By:

ImageAnalyzer and VideoAnalyzer



Detailed Description

The AbstractImageAnalyzer component utilizes on-device deep learning models to classify images and detect objects within images.

The following code snippet shows the component used for performance tuning. Information relating to the delegates that were used in the inference , as well as the time it took to perform the inference, are returned.

ImageAnalyzer {
    properties: {
        "use_gpu": true, // enable GPU delegate
        "use_nnapi": false, // disable NNAPI delegate
        "verbose": 1
    onFinished: {
        // for example, { "profile": { "delegates": [], "inference_time": 23 }, "use_gpu": true, "use_nnapi": false, "verbose": 1 }

This component is in beta. Elements are subject to change in future releases.


Status enumeration

Enum describing potential statuses of the image analyzer. Informs the status property.


Property Documentation

acceleration : bool

If set to true, the analyzer will try to use the Android Neural Networks API (NNAPI) delegate on an Android device to achieve hardware acceleration. No other forms of hardware acceleration delegates are supported. Defaults to true.

classNames : QStringList

The list of class names that are used to label analysis results.

[read-only] error : DeepLearningError

Describes any error encountered by the image analyzer.

minimumScore : double

The minimum score for the analyzer to achieve before providing a result. Default value is 0.5.

modelSource : url

The source of the deep learning model file.

numThreads : int

The number of threads that the image analyzer is running on.

properties : object

Parameters for benchmarking or performance tuning.

The verbose parameter is used to retrieve performance metrics after image analysis completes.

For more parameters, see https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/tools/benchmark/README.md.

[read-only] status : Status

The current status of the image analyzer. Informed by the Status enum.

Method Documentation

color classColor(string name)

Returns the color corresponding to the class name.

The name parameter

The name of the class to return the color of.

See also setClassColor().

setClassColor(string name, color color)

Sets the color for the specified class name.

The name parameter

The name of the class to set the color of.

The color parameter

The color to set the class to.

See also classColor().

Your browser is no longer supported. Please upgrade your browser for the best experience. See our browser deprecation post for more details.