AI-102: How to Extract Text from Scanned Documents Using Azure AI Vision?

Learn how to utilize Azure AI Vision’s Read API for OCR and text extraction from images. Find out the best practices to streamline your application development.

Question

Your organization, Nutex Inc., is developing an application that requires extracting text from scanned documents. To achieve this, you plan to use Azure AI Vision’s Read API.

Which visual feature should you specify to make an optical character recognition (OCR) request to the ImageAnalysis function?

A. VisualFeatures.Tags
B. VisualFeatures.Objects
C. VisualFeatures.Read
D. VisualFeatures.Caption

Answer

C. VisualFeatures.Read

Explanation

You would specify VisualFeatures.Read as the visual feature to make an OCR request to the ImageAnalysis function.

When using C# for developing the application, you can use the below code snippet:

ImageAnalysisResult result = client.Analyze(
<image-to-analyze>,
VisualFeatures.Read);

When using Python for developing the application, you can use the below code snippet:

result = client.analyze(
image_url=<image_to_analyze>,
visual_features=[VisualFeatures.READ]
)

The VisualFeatures.Read feature is specifically designed for OCR tasks. When you specify this feature in your request to ImageAnalysis, the Azure AI Vision service focuses on recognizing and extracting text from the image. This feature is essential for applications that need to process text from scanned documents, handwritten notes, receipts, and other image-based text sources.

You would not specify the VisualFeatures.Tags visual feature to make an OCR request to the ImageAnalysis function. The VisualFeatures.Tags feature identifies and labels various elements within an image, such as objects, scenery, settings, and actions. This feature is useful for general image analysis and categorization. However, it is not designed for text extraction or OCR tasks.

You would not specify the VisualFeatures.Objects visual feature to make an OCR request to the ImageAnalysis function. The VisualFeatures.Objects feature returns the bounding box for each detected object within an image. This feature is useful for detecting and locating objects, such as people, vehicles, or furniture, within a picture. However, it is not intended for OCR purposes and does not extract or recognize text from images.

You would not specify the VisualFeatures.Caption visual feature to make an OCR request to the ImageAnalysis function. The VisualFeatures.Caption feature generates a natural language description of the image, summarizing the content of the image in a short sentence. This feature is useful for generating a caption for an image, making it easier to understand the overall content of an image without extracting specific text.

Microsoft Azure AI Engineer Associate AI-102 certification exam practice question and answer (Q&A) dump with detail explanation and reference available free, helpful to pass the Microsoft Azure AI Engineer Associate AI-102 exam and earn Microsoft Azure AI Engineer Associate AI-102 certification.