Method: projects.locations.productSets.import

Asynchronous API that imports a list of reference images to specified product sets based on a list of image information.

The google.longrunning.Operation API can be used to keep track of the progress and results of the request. Operation.metadata contains BatchOperationMetadata. (progress) Operation.response contains ImportProductSetsResponse. (results)

The input source of this method is a csv file on Google Cloud Storage. For the format of the csv file please see ImportProductSetsGcsSource.csv_file_uri.

HTTP request


The URL uses Google API HTTP annotation syntax.

Path parameters



The project in which the ProductSets should be imported.

Format is projects/PROJECT_ID/locations/LOC_ID.

Request body

The request body contains data with the following structure:

JSON representation
  "inputConfig": {


The input content for the list of requests.

Response body

If successful, the response body contains an instance of Operation.

Authorization Scopes

Requires one of the following OAuth scopes:


For more information, see the Authentication Overview.


The input content for the productSets.import method.

JSON representation
  "gcsSource": {


The Google Cloud Storage location for a csv file which preserves a list of ImportProductSetRequests in each line.


The Google Cloud Storage location for a csv file which preserves a list of ImportProductSetRequests in each line.

JSON representation
  "csvFileUri": string


The Google Cloud Storage URI of the input csv file.

The URI must start with gs://.

The format of the input csv file should be one image per line. In each line, there are 8 columns.

  1. image-uri
  2. image-id
  3. product-set-id
  4. product-id
  5. product-category
  6. product-display-name
  7. labels
  8. bounding-poly

The image-uri, product-set-id, product-id, and product-category columns are required. All other columns are optional.

If the ProductSet or Product specified by the product-set-id and product-id values does not exist, then the system will create a new ProductSet or Product for the image. In this case, the product-display-name column refers to displayName, the product-category column refers to productCategory, and the labels column refers to productLabels.

The image-id column is optional but must be unique if provided. If it is empty, the system will automatically assign a unique id to the image.

The product-display-name column is optional. If it is empty, the system sets the displayName field for the product to a space (" "). You can update the displayName later by using the API.

If a Product with the specified product-id already exists, then the system ignores the product-display-name, product-category, and labels columns.

The labels column (optional) is a line containing a list of comma-separated key-value pairs, in the following format:


The bounding-poly column (optional) identifies one region of interest from the image in the same manner as referenceImages.create. If you do not specify the bounding-poly column, then the system will try to detect regions of interest automatically.

At most one bounding-poly column is allowed per line. If the image contains multiple regions of interest, add a line to the CSV file that includes the same product information, and the bounding-poly values for each region of interest.

The bounding-poly column must contain an even number of comma-separated numbers, in the format "p1_x,p1_y,p2_x,p2_y,...,pn_x,pn_y". Use non-negative integers for absolute bounding polygons, and float values in [0, 1] for normalized bounding polygons.

The system will resize the image if the image resolution is too large to process (larger than 20MP).

Try it!

Was this page helpful? Let us know how we did:

Send feedback about...

Cloud Vision API
Need help? Visit our support page.