Microsoft Certified: Azure AI Engineer AssociateImplement image and video processing solutionsHard
A local government agency needs to process a large volume of scanned historical documents, many of which are in different languages (e.g., English, Spanish, French) and contain both printed and handwritten text. They require not only text extraction but also automatic identification of the language for each text block. Which Azure Computer Vision API, combined with which other Azure AI service, provides the most comprehensive solution?
- AOCR (Legacy) API + Azure Translator
- BAnalyze Image + Azure Translator
- CRead API + Text Analytics Language Detection
- DForm Recognizer + Text Analytics Entity Recognition
Show answer & explanationAnswer & explanation
Correct answer: C. Read API + Text Analytics Language Detection
The Read API is the most capable OCR for mixed printed/handwritten text and multiple languages. While Read API can detect some languages, combining it with Text Analytics Language Detection provides more robust and granular language identification for each extracted text block, especially for diverse historical documents.
Why the other options are wrong
- A. Legacy OCR is inferior for complex documents; Azure Translator is for translation, not language detection of source text.
- B. Analyze Image does not extract detailed text; Azure Translator is for translation.
- D. Form Recognizer is for structured forms; Text Analytics Entity Recognition identifies specific entities, not the overall language of a block.
Computer Vision Read API + Text Analytics Language Detection
A powerful combination for extracting text from complex, multi-language documents using Read API, followed by precise language identification of each extracted text block using Text Analytics.
- Read API handles printed, handwritten, and mixed text with high accuracy.
- Read API provides bounding boxes and structural information.
- Text Analytics Language Detection accurately identifies the language of short or long text inputs.
- Ideal for digitizing diverse historical archives.
Memory trick: Read API pulls the words out, Text Analytics tells you what language they speak.