Published at: 2025-10-30

OCR Text Recognition

1. Definition of OCR

OCR (Optical Character Recognition) is a text recognition technology that extracts textual information from various media (paper, images, documents, billboards, etc.).

2. Business Scenarios

Type Business Scenario
ID & Document Recognition ID Card Recognition: Automatically captures and records user identity information. Applicable in finance, insurance, e-commerce, O2O, and Live scenarios for实名认证 (real-name authentication), reducing input costs and mitigating business risks.
ID & Document Recognition Bank Card Recognition: Supports remote identity verification in financial scenarios, enhancing user experience by minimizing manual input.
ID & Document Recognition Business License Recognition: Streamlines merchant onboarding for e-commerce, retail, and O2O industries by automating资质审查 (qualification review), improving operational efficiency.
ID & Document Recognition Household Register Recognition: Extracts details like name, gender, birthplace, date of birth, and ID number. Used in新生儿建档 (newborn records),户口迁移 (household relocation), and government services to expedite verification.
ID & Document Recognition Passport Recognition: Facilitates automatic data entry for international travel bookings and hotel check-ins, boosting efficiency and user satisfaction.
Financial Document Recognition Invoice Recognition: Automates expense reimbursement by splitting and structuring receipts. Ideal for internal accounting, reducing manual workload and simplifying processes.
Transportation Recognition Vehicle License Recognition: Combines with driver’s license and ID recognition for ride-hailing or truck driver registration, optimizing user onboarding.
General Use Digitizes paper documents to enhance data entry, archiving, and retrieval efficiency.

3. Configuration Steps

  1. Identify the type of information to recognize and purchase the corresponding resource package in CRM.

image

  1. Create fields to store recognized data (e.g., name and ID number for ID cards).

image

  1. After the resource package is activated:
    • Navigate to the object and create an Images or Files field (or use an existing one).
    • Enable OCR → Select recognition type (e.g., ID card) → Configure field mapping.

image
image

  1. Auto-fill: Upload an ID card image on the object’s creation page, and the system populates fields like name and ID number automatically.

image
Example of OCR recognition for images (attachment-based OCR works similarly).

4. FAQs

4.1 Supported Formats

  • Images: PNG, JPG, JPEG, BMP, TIFF, PNM, WebP
  • Invoices: PDF and OFD formats.

4.2 Troubleshooting

Common issues affecting OCR accuracy:
1. Image size <15px.
2. Poor quality (e.g., low brightness).
3. Obstructions like watermarks, stamps, or folds.
4. Mismatched document type (e.g., passport submitted for ID card recognition).
5. Refer to error codes if returned.

Error Code Error Message Description
1 Unknown error Retry or contact support.
2 Service temporarily unavailable Retry later.
3 Unsupported openapi method Check URL for non-English characters.
4 Open api request limit reached Cluster quota exceeded. Retry.
6 No permission to access data Enable OCR APIs in app settings.
… … … (Full table retained)

Persistent errors? Contact the product team.

Submit Feedback
Previous
Date Field
Next
Positioning Fields