Introduction

Taggun's API offers advanced fraud detection tools to enhance the security of your customer engagement promotions and employee reimbursement processes. Our tampering detection feature helps identify fraudulent attempts to modify receipt images before submission on your platform.

Common Fraud Cases Handled

Photoshopped Receipts: Detect receipts altered using Photoshop or other image editing software.
File Metadata Manipulation: Identify image and PDF files with suspicious metadata indicating tampering.
Copy-Paste Image Fraud: Catch attempts to submit receipts where parts of the image have been copied and pasted within the same image.

How to Set Up

Contact Taggun: Reach out to [email protected] to enable this feature for your account.
Use verbose endpoints: The fraud detection feature is only available in all our verbose endpoints.

Understanding the response properties

Field Name	Type	Description
entities.tamperingDetection	array	An object containing the tampering detection results.
entities.tamperingDetection.isTampered	boolean	Indicates whether the receipt has been tampered with.
entities.tamperingDetection.tamperedScore	number	The calculated score indicating the likelihood of tampering, with 0 indicating a genuine receipt and 0.99 indicating a tampered receipt
entities.tamperingDetection.details	object	Additional details regarding the tampering detection analysis..

How does it work?

The submitted file undergoes several tampering detection analyses, including:

Error Level Analysis (ELA): Analyzes different compression levels within an image to identify discrepancies indicative of tampering
Metadata Check: Examines the image's metadata for inconsistencies suggesting manipulation.
Copy and Paste Detection: Identifies areas within the image where parts have been duplicated, indicating potential fraudulent alterations.

If any of these methods detect tampering, the receipt will be flagged for tampering and it is recommended for manual review or rejection.

Examples

A malicious user modifies the total amount using Photoshop.

Photoshopped Receipt

RECEIPT & INVOICE OCR API

An example response when a tampered receipt is submitted

{
  ...
  "entities": {
    "tamperDetection": {
      "data": {
        "isTampered": true,
        "tamperedScore": 0.82,
        "details": {
          "elaTampered": false,
          "metadataTampered": true,
          "copyPasteTampered": false,
          "duplicates": [],
          "normalisedOverallMeanDiff": 0.013531381881449533
        }
      }
  }
  ...
}

Start Building

Contact Taggun now [email protected] to get access to this feature, and then dive straight into the API reference to start building fraud-proof applications!