E-Paper Pro

Price Reduced! Auto Tagging USD 20/mo

The best E-Paper
CMS Solution
Menu
Live Demo Or
Client Area

AutoTag API — Documentation

Last updated: 2025-01-31

A simple REST endpoint that receives a page image and returns layout sections detected on that page (bounding boxes, class, confidence). Ideal for e-paper / newspaper page processing, automated tagging and layout-aware rendering.

Overview

The AutoTag API analyzes a page image and returns an array of detected layout sections. Each detection contains the bounding box (xmin, ymin, xmax, ymax), a confidence score, a numeric class id, and a human-readable name (for example advertisement, headline, article, etc.).

Use cases:

  • Automatic region extraction for e‑paper conversions
  • Tagging ads vs editorial content
  • Feeding region boxes to OCR engines selectively

Endpoint

POST https://myaccount.epaperpro.com/wp-json/autotag/v1/process

Content-Type: application/json

Request payload

{
  "image_url": "https://epaper.roznamasahara.com/imgsrc/108439/1/1681-2469-2679-3134",
  "license_key": "67f368baa1"
}
  • image_url (string) — publicly reachable URL of the page image (jpg/png/webp). The service fetches this URL to process the image.
  • license_key (string) — your API license key for authentication/usage tracking.

Note: If your image URL requires a header or authentication, host the image on a publicly reachable URL (S3, CDN) or contact support for private ingestion options.

Example success response

{
    "result": [
        {
            "xmin": 254.4713439941,
            "ymin": 176.5631713867,
            "xmax": 984.3304443359,
            "ymax": 621.6746826172,
            "confidence": 0.7537087798,
            "class": 0,
            "name": "advertisement"
        }
    ]
}

Fields explained:

  • xmin, ymin, xmax, ymax — coordinates in pixels relative to the top-left of the image.
  • confidence — floating number (0.0 to 1.0) showing model confidence for that detection.
  • class — numeric id for detected label (internal mapping).
  • name — human-readable class label.

Example usage

cURL

curl -X POST "https://myaccount.epaperpro.com/wp-json/autotag/v1/process" \
  -H "Content-Type: application/json" \
  -d '{
    "image_url": "https://epaper.roznamasahara.com/imgsrc/108439/1/1681-2469-2679-3134",
    "license_key": "67f368baa"
  }'

JavaScript (Fetch)

async function autoTag(imageUrl, licenseKey) {
  const res = await fetch('https://myaccount.epaperpro.com/wp-json/autotag/v1/process', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({ image_url: imageUrl, license_key: licenseKey })
  });
  if (!res.ok) throw new Error('API error ' + res.status);
  return res.json();
}

// usage
autoTag('https://epaper.roznamasahara.com/imgsrc/108439/1/1681-2469-2679-3134', '67f368ba1627')
  .then(console.log)
  .catch(console.error);

Python (requests)

import requests

url = 'https://myaccount.epaperpro.com/wp-json/autotag/v1/process'
payload = {
    'image_url': 'https://epaper.roznamasahara.com/imgsrc/108439/1/1681-2469-2679-3134',
    'license_key': '67f36a1627'
}
res = requests.post(url, json=payload)
res.raise_for_status()
print(res.json())

PHP (curl)

$ch = curl_init('https://myaccount.epaperpro.com/wp-json/autotag/v1/process');
$data = json_encode([
  'image_url' => 'https://epaper.roznamasahara.com/imgsrc/108439/1/1681-2469-2679-3134',
  'license_key' => '67f368b162'
]);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, 'POST');
curl_setopt($ch, CURLOPT_POSTFIELDS, $data);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, [
  'Content-Type: application/json',
  'Content-Length: ' . strlen($data)
]);

$response = curl_exec($ch);
curl_close($ch);

echo $response;

Example page (HTML) — quick demo

<!doctype html>
<html>
<head>
  <meta charset="utf-8">
  <title>AutoTag Demo</title>
  <style>
    body{font-family:system-ui,Arial;max-width:900px;margin:24px auto}
    img{max-width:100%}
    pre{background:#f6f8fa;padding:12px;border-radius:6px}
  </style>
</head>
<body>
  <h1>AutoTag API — Demo</h1>
  <p>Paste an image URL and click <strong>Process</strong>.</p>
  <input id="imageUrl" value="https://epaper.roznamasahara.com/imgsrc/108439/1/1681-2469-2679-3134" style="width:100%" />
  <button id="go">Process</button>
  <div id="out"></div>

  <script>
    document.getElementById('go').addEventListener('click', async () => {
      const img = document.getElementById('imageUrl').value.trim();
      const payload = { image_url: img, license_key: '67f368162' };
      const res = await fetch('https://myaccount.epaperpro.com/wp-json/autotag/v1/process', {
        method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(payload)
      });
      const json = await res.json();
      document.getElementById('out').innerHTML = '<h3>Result</h3><pre>'+JSON.stringify(json, null, 2)+'</pre>';
    });
  </script>
</body>
</html>

This page displays raw JSON — for production you would use the bounding boxes to draw overlays on the image and create clickable regions.

How to interpret coordinates and draw boxes

  1. Load the original image into an HTML <img> element.
  2. Get the rendered width/height and the original image natural width/height.
  3. The API returns pixel coordinates relative to the original image (natural width/height). If you render the image scaled, compute a scale factor and multiply the returned coordinates by that factor to draw accurate overlays.

Example (scale):

scaleX = renderedWidth / naturalWidth
scaleY = renderedHeight / naturalHeight
screenX = xmin * scaleX
screenY = ymin * scaleY
width = (xmax - xmin) * scaleX
height = (ymax - ymin) * scaleY

Error responses

  • 400 Bad Request — malformed JSON or missing fields.
  • 401 Unauthorized — invalid or missing license_key.
  • 422 Unprocessable Entity — image URL not reachable or unsupported format.
  • 500 Internal Server Error — processing failure (retry with exponential backoff).

Always check the HTTP status code before parsing JSON.

Rate limiting & quotas

Your license key is subject to per-account quotas and rate limits. If you hit limits, you’ll receive HTTP 429 Too Many Requests with a Retry-After header. Contact support to increase quota.

Best practices

  • Use CDN-hosted images for lower fetch latency.
  • Cache results for identical image_url requests to avoid re-processing the same image.
  • Filter detections by confidence (e.g., only accept detections > 0.5).
  • If you need absolute layout precision, process images at their original resolution.

Security

  • Keep license_key secret. Do not embed it in client-side code for production-facing applications. The demo above is for quick testing only. For production, call the AutoTag API from your server or implement a token exchange.

Changelog

  • 2025-12-06 — Initial documentation created.

If you’d like, I can also add a ready-made client library snippet, an overlay visualizer (HTML + canvas) that draws boxes on the image, or convert this doc into a printable Markdown/HTML page.