org.apache.beam.sdk.extensions.ml

Class CloudVision.AnnotateImagesFromBytesWithContext

java.lang.Object

org.apache.beam.sdk.transforms.PTransform<PCollection<T>,PCollection<java.util.List<com.google.cloud.vision.v1.AnnotateImageResponse>>>

org.apache.beam.sdk.extensions.ml.CloudVision.AnnotateImagesFromBytesWithContext

All Implemented Interfaces:

java.io.Serializable, HasDisplayData

Enclosing class:

CloudVision

public static class CloudVision.AnnotateImagesFromBytesWithContext
extends PTransform<PCollection<T>,PCollection<java.util.List<com.google.cloud.vision.v1.AnnotateImageResponse>>>

Accepts KVs of ByteString (encoded image contents) and ImageContext. It's possible to add ImageContext to each image to be annotated.

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	org.apache.beam.sdk.extensions.ml.AnnotateImages.PerformImageAnnotation The DoFn performing the calls to Cloud Vision API.

Field Summary

Fields

Modifier and Type	Field and Description
protected PCollectionView<java.util.Map<T,com.google.cloud.vision.v1.ImageContext>>	contextSideInput
protected int	desiredRequestParallelism
protected java.util.List<com.google.cloud.vision.v1.Feature>	featureList

Fields inherited from class org.apache.beam.sdk.transforms.PTransform

name, resourceHints

Constructor Summary

Constructors

Constructor and Description
AnnotateImagesFromBytesWithContext(java.util.List<com.google.cloud.vision.v1.Feature> featureList, long batchSize, int desiredRequestParallelism)

Method Summary

Modifier and Type	Method and Description
PCollection<java.util.List<com.google.cloud.vision.v1.AnnotateImageResponse>>	expand(PCollection<T> input) Applies all necessary transforms to call the Vision API.
com.google.cloud.vision.v1.AnnotateImageRequest	mapToRequest(KV<ByteString,com.google.cloud.vision.v1.ImageContext> input, @Nullable com.google.cloud.vision.v1.ImageContext ctx) Maps KV of ByteString (encoded image contents) and ImageContext to AnnotateImageRequest.

Methods inherited from class org.apache.beam.sdk.transforms.PTransform

compose, compose, getAdditionalInputs, getDefaultOutputCoder, getDefaultOutputCoder, getDefaultOutputCoder, getKindString, getName, getResourceHints, populateDisplayData, setResourceHints, toString, validate, validate

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

contextSideInput

protected final PCollectionView<java.util.Map<T,com.google.cloud.vision.v1.ImageContext>> contextSideInput

featureList

protected final java.util.List<com.google.cloud.vision.v1.Feature> featureList

desiredRequestParallelism

protected final int desiredRequestParallelism

Constructor Detail

AnnotateImagesFromBytesWithContext

public AnnotateImagesFromBytesWithContext(java.util.List<com.google.cloud.vision.v1.Feature> featureList,
                                          long batchSize,
                                          int desiredRequestParallelism)

Method Detail

mapToRequest

public com.google.cloud.vision.v1.AnnotateImageRequest mapToRequest(KV<ByteString,com.google.cloud.vision.v1.ImageContext> input,
                                                                    @Nullable com.google.cloud.vision.v1.ImageContext ctx)

Maps KV of ByteString (encoded image contents) and ImageContext to AnnotateImageRequest.

Parameters:

input - Input element.

ctx - optional image context.

Returns:

valid request element.

expand

public PCollection<java.util.List<com.google.cloud.vision.v1.AnnotateImageResponse>> expand(PCollection<T> input)

Applies all necessary transforms to call the Vision API. In order to group requests into batches, we assign keys to the requests, as GroupIntoBatches works only on KVs.

Specified by:

expand in class PTransform<PCollection<T>,PCollection<java.util.List<com.google.cloud.vision.v1.AnnotateImageResponse>>>