gcloud alpha ml vision product-search images create

gcloud alpha ml vision product-search images create - create a reference image
gcloud alpha ml vision product-search images create (REFERENCE_IMAGE : --location=LOCATION --product=PRODUCT) --storage-location=STORAGE_LOCATION [--bounding-polygon=[normalized-vertices=NORMALIZED-VERTICES],[vertices=VERTICES]] [GCLOUD_WIDE_FLAG]
(ALPHA) create a reference image.
Reference image resource - Reference image to create. The arguments in this group can be used to specify the attributes of this resource. (NOTE) Some attributes are not given arguments in this group but can be set in other ways. To set the [project] attribute: provide the argument [reference_image] on the command line with a fully specified name; set the property [core/project]; provide the argument [--project] on the command line. This must be specified.
ID of the reference image or fully qualified identifier for the reference image. This positional must be specified if any of the other arguments in this group are specified.
The location of reference image.
The ID of product.
The Google Cloud Storage URI of the reference image. The URI must start with 'gs://'.
Bounding polygon around the areas of interest in the reference image. If this field is empty, the system will try to detect regions of interest. This flag is repeatable to specify multiple bounding polygons. At most 10 bounding polygons will be used.

A bounding polygon can be specified by a list of vertices or normalized vertices or both. A vertex (x, y) represents a 2D point in the image. x, y are integers and are in the same scale as the original image. The normalized vertex coordinates are relative to original image and range from 0 to 1.

Because of the complexity of this flag, it should be specified with the --flags-file. See $ gcloud topic flags-file for details. See the examples section for how to use --bounding-polygon in --flags-file.

These flags are available to all commands: --account, --billing-project, --configuration, --flags-file, --flatten, --format, --help, --impersonate-service-account, --log-http, --project, --quiet, --trace-token, --user-output-enabled, --verbosity.

Run $ gcloud help for details.

This command uses the vision/v1 API. The full documentation for this API can be found at: https://cloud.google.com/vision/
To create a reference image 'my-image' in product 'my-product' in the location 'us-east1' from image of 'gs://my-bucket/image.jpg' with two bounding polygons, run:
gcloud alpha ml vision product-search images create my-image  --location=my-location --product=my-product  --storage-location=gs://my-bucket/image.jpg  --flags-file=my-flags.yaml

The content of 'my-flags.yaml' is as following:

  - --bounding-polygon:
        - {x: 0, y: 0}
        - {x: 0, y: 10}
        - {x: 10, y: 0}
        - {x: 10, y: 10}
  - --bounding-polygon:
        - {x: 0.8, y: 0.8}
        - {x: 0.8, y: 1}
        - {x: 1, y: 0.8}
        - {x: 1, y: 1}
This command is currently in ALPHA and may change without notice. If this command fails with API permission errors despite specifying the right project, you may be trying to access an API with an invitation-only early access allowlist. This variant is also available:
gcloud beta ml vision product-search images create