Tutorials  

Looking for a powerful way to analyze images using artificial intelligence? In this step-by-step tutorial, we’ll explore how to set up and use the Google Cloud Vision API with a simple Node.js application. Our example will focus on determining whether a product image is likely to appeal to children by analyzing labels and dominant colors. Whether you’re a developer, marketer, or just curious about AI, this guide has you covered.

Introduction

Google Cloud Vision API is a powerful tool that uses AI to extract information from images. You can detect objects, faces, text, logos, and more, making it perfect for everything from marketing campaigns to compliance checks. In our example, we’ll use it to classify product images as “child-friendly” or “not particularly child-friendly” based on detected labels and bright, vibrant colors.

Step 1: Set Up Your Google Cloud Project

  1. Create or select a project: Visit console.cloud.google.com, then click “Create Project” or choose an existing project.
  2. Enable the API: In the left-hand menu, select APIs & Services > Library. Search for Vision API and click Enable.

Step 2: Configure a Service Account and Download the Key

  1. Go to IAM & Admin, then select Service Accounts.
  2. Click “Create Service Account,” provide a name (e.g., children-product-packaging-analysis), and assign the Editor role.
  3. Under “Manage Keys,” click “Add Key,” then “Create new key,” select JSON, and download the generated JSON file. Keep this file secure!

Step 3: Set Up Your Node.js Project

  1. Create a folder: For example, “google-cloud-vision-api.”
  2. Place the key file inside: Copy the downloaded JSON key file to your project folder.
  3. Initialize Node.js: In your project folder, run:
    npm init -y
  4. Install the Vision library:
    npm install @google-cloud/vision

Step 4: Create the index.js File

Inside your project folder, create a file named index.js. Here’s a basic example of the code you can use to analyze images for child-friendliness. Make sure to replace PATH_TO_JSON_KEY with your actual JSON key filename, and YOUR_IMAGE_PATH with the path to the image you want to analyze: 

// index.js

// 1. Import and configure the Vision client
const vision = require('@google-cloud/vision');

// Create a new client, passing in your service account key file
const client = new vision.ImageAnnotatorClient({
  keyFilename: 'PATH_TO_JSON_KEY.json'
});

// 2. Define an async function to analyze child-friendliness
async function analyzeImage(imagePath) {
  try {
    // Detect labels in the image
    const [labelResult] = await client.labelDetection(imagePath);
    const labels = labelResult.labelAnnotations.map(label => label.description.toLowerCase());
    
    // Detect image properties for dominant color analysis
    const [propsResult] = await client.imageProperties(imagePath);
    const colors = propsResult.imagePropertiesAnnotation.dominantColors.colors;
    
    // Set thresholds for child-friendly labels and color vibrancy
    const childFriendlyLabels = ['toy', 'cartoon', 'game', 'animal', 'fun', 'colorful'];
    const colorThreshold = 0.5;       // Adjust for brightness
    const pixelFractionThreshold = 0.3; // Adjust fraction of the image that must be vibrant
    
    // Check if any labels match child-friendly keywords
    const hasChildFriendlyLabels = labels.some(label => childFriendlyLabels.includes(label));
    
    // Check if any dominant color meets brightness and pixel fraction requirements
    const hasBrightColors = colors.some(color => {
      return color.color.red / 255 > colorThreshold ||
             color.color.green / 255 > colorThreshold ||
             color.color.blue / 255 > colorThreshold &&
             color.pixelFraction > pixelFractionThreshold;
    });
    
    // Determine if the image is child-friendly
    const isChildFriendly = hasChildFriendlyLabels || hasBrightColors;
    
    if (isChildFriendly) {
      console.log('This image is likely attractive to children.');
      return true;
    } else {
      console.log('This image is not particularly attractive to children.');
      return false;
    }
    
  } catch (err) {
    console.error('Error analyzing image:', err);
    return null;
  }
}

// 3. Test the function with your chosen image
const testImagePath = 'YOUR_IMAGE_PATH';
analyzeImage(testImagePath);

Step 5: Test Your Analysis

  1. Open your terminal and navigate to your project folder.
  2. Run:
    node index.js
  3. Check the console output to see if the image is “likely attractive to children” or “not particularly attractive to children.”
  4. Adjust parameters (colorThreshold, pixelFractionThreshold, or label keywords) if you need more accurate results.

Step 6: Additional Tips and Resources

  • Github Repository: https://github.com/codingmoney/google-cloud-vision-api-tutorial
  • Try different images to see how the analysis changes based on labels and color properties.
  • Use a GitHub repository to store your code and instructions for easy collaboration.
  • Consider expanding to other Vision API features like text detection, logo detection, or face detection to enhance your application.

Explore monetization ideas

AI image analysis can help e-commerce sites, compliance checks, targeted advertising, and more. Checkout 10 Ways to Make Money with AI Vision for more ideas.

Conclusion

Congratulations! You’ve successfully set up a Node.js application using Google Cloud Vision API to analyze images for child-friendliness. Whether you’re creating a marketing campaign or building a platform to categorize product images, this AI-driven approach can save time and offer deeper insights. Don’t forget to subscribe to newsletter and share if you found this tutorial useful. Happy coding!