Microsoft Certified: Azure AI Engineer AssociateImplement image and video processing solutionsMedium
A logistics company wants to automate the processing of shipping labels. These labels often contain printed text in various fonts, handwritten notes, and sometimes faded or skewed text. They need to extract all text content accurately, regardless of its orientation or quality. Which Azure Computer Vision API should they use?
- AAnalyze Image (Tags)
- BRead API
- COCR (Optical Character Recognition) API (Legacy)
- DDescribe Image API
Show answer & explanationAnswer & explanation
Correct answer: B. Read API
The Read API in Azure Computer Vision is specifically designed for robust OCR of high-resolution images, including documents and labels with mixed fonts, handwritten text, and varying orientations, offering superior accuracy over the legacy OCR API.
Why the other options are wrong
- A. Analyze Image (Tags) generates descriptive tags, not text extraction.
- C. The legacy OCR API is less accurate and capable than the Read API, especially for mixed or handwritten text.
- D. Describe Image API generates a human-readable sentence describing image content, not text extraction.
Computer Vision Read API
An advanced OCR capability within Azure Computer Vision designed for high-accuracy text extraction from a wide range of images, including documents and labels with mixed content.
- Supports printed and handwritten text
- Handles various fonts, orientations, and image qualities
- Provides text lines and word bounding box locations
- Offers multi-language support
Memory trick: The Read API is the 'smartest reader' for any text on a picture.