Can't change MaxResults on Cloud Vision Labeling

Question

I'm using Google Vision to detect photos on an image and the results are limited to the top 10. I want to extend this number to 50. How do I do that?

Here is my code:

const client = new vision.ImageAnnotatorClient();

  for (const file of allFiles) {
    if (file.metadata.contentType.startsWith("image/")) {
      const [resultImageLabel] = await client.labelDetection(
        `gs://${bucketName}/${file.name}`
      );

      const labels = resultImageLabel.labelAnnotations;
      const faceDetection = resultFaceDetection.faceAnnotations;

      const labelArray = [];
      const labelScore = {};

      // image Labelling
      for (const label of labels) {
        labelArray.push(label.description);
        labelScore[label.description] = label.score;
      }

      console.log(labelArray);
      console.log(labelScore);
    }
  }

The answers that i've seen are in JSON format so i added at the beginning of the for loop:

const request = {
      image: {
        source: { imageUri: `gs://${bucketName}/${file.name}` },
      },
      features: [
        {
          maxResults: 50,
        },
      ],
    };

// Changed from image path to request
const [resultImageLabel] = await client.labelDetection(request);

Then I get the following error:

functions: Error: Setting explicit features is not supported on this method. Use the #annotateImage method instead.

Looked through the "#annotateImage" documentation and I don't see what I'm missing.

When I comment out the "features" section of the code, everything works as before.

This is running in an https Firebase Cloud Function.

Can you try `await client.annotateImage` instead of `await client.labelDetection` in your code? Please let me know if this is helpful or not. — kiran mathew, Aug 23 '23 at 05:09

score 1 · Accepted Answer · answered Aug 23 '23 at 07:32

For your requirements,you can consider the following code below as an example:

const client = new vision.ImageAnnotatorClient();
async function testing(){

const allFiles = [
{ name: 'gs://bucket_name/file1.jpeg', metadata: { contentType: 'image/jpeg' } },
];

for (const file of allFiles) {
if (file.metadata.contentType.startsWith("image/")) {
const request = {
image: {
source: { imageUri: file.name },
},
features: [
{
type: 'LABEL_DETECTION',
maxResults: 50,
},
],
};

const [resultImageLabel] = await client.annotateImage(request);
const labels = resultImageLabel.labelAnnotations;
const labelArray = [];
const labelScore = {};

for (const label of labels) {
labelArray.push(label.description);
labelScore[label.description] = label.score;
}
console.log(labelArray);
console.log(labelScore);
}
} 
} 

testing();

Sample image used :https://unsplash.com/photos/wfwUpfVqrKU

From the error message it is clear that `features’ is not supported on the labelDetection method so instead that method we have to use the ‘annotateImage’ method.You can find more information from these link1 and link2.

Can't change MaxResults on Cloud Vision Labeling

1 Answers1