Basic Information
| Field | Details |
|---|---|
| English name | Amazon Textract |
| Full name | Amazon Textract |
| Chinese description | 文档文字与结构提取 |
| Japanese description | 文書データ抽出 |
| Exam frequency | ⭐⭐ |
| Often confused with | Rekognition |
In one sentence
Extract text, tables, form fields, and structured data from scanned documents.
Key points
- It fits invoices, receipts, applications, contracts, and identity documents.
- Compared with basic OCR, Textract focuses on key-value pairs, tables, and document structure.
- For objects and scenes in general images or video, compare Rekognition.
Exam focus
- Choose Textract for scanned documents, tables, forms, and field extraction.
Common pitfalls
- Do not choose by product name alone; confirm data type, latency, control, operations, and cost constraints.
Remember
Textract extracts document structure; Rekognition sees image content.