Understanding the PicDefense API: A Comprehensive Breakdown
The PicDefense API allows you to automate the Risk Forensics pillar of our platform. By hitting our /api/v2/ endpoints, you trigger a multi-engine analysis that leverages TinEye, Google Cloud Vision, and our proprietary OpenAI risk scoring.
This guide breaks down the JSON payload returned by our API so you can programmatically route flagged images into your Proof Vault for manual compliance verification.
Reverse Search Data (Powered by TinEye)
Our backend integrates directly with the TinEye API for reverse image search functionality.
Backlink Count & Detail
Returns the exact URLs where the image was found across the web. If an asset matches known stock libraries (like Getty or Shutterstock), this data serves as a critical source hint for reconstructing provenance.
Image Similarity
A score (0 to 100) indicating how much your file has been modified from the source.
Machine Vision Data (Powered by GCP Vision)
We pipe the image through the Google Cloud Vision API to detect protected elements.
Face, Landmark, & Logo Detection
Identifies elements that carry higher trademark or privacy release requirements.
EXIF Data Analysis
Extracts deep metadata, including camera make, original dimensions, and embedded copyright tags.
The Forensic Risk Score (API field: picRisk)
This is our synthesized triage metric. The endpoint returns a calculated Forensic Risk Score with a risk level (High, Medium, Low) based on the combined forensic data.
Architectural Guidance: How to Use This Data
Remember our core thesis: Risk helps you triage; Proof closes the loop. Do not use the API as a final legal verdict. Instead, build your integration to automatically flag any image with a “High” Forensic Risk Score and push it into a review queue where your team can attach the required license in the Proof Vault.
For detailed request/response schemas, type definitions, and endpoint routing, review our complete Swagger documentation at https://app.picdefense.io/apidocs/.