AI-102 Question 123
Select 4You are building an application that processes scanned receipts and extracts text to analyze purchase patterns. The application uses Azure AI Vision to extract text from images. Which of the following steps are required to successfully implement this solution?
- A
Create an Azure Computer Vision resource and retrieve its endpoint and key.
- B
Use the Azure AI Vision Optical Character Recognition (OCR) API to analyze the receipts.
- C
Train a custom model using Azure Custom Vision for text extraction.
- D
Parse the JSON response returned by the OCR API to extract the relevant text.
- E
Enable the 'Read' API mode for analyzing handwritten and printed text.
Show answer and explanation
Correct answers: A, B, D, E
Explanation
To extract text from images, you need to create an Azure Computer Vision resource and use the OCR API with the appropriate 'Read' mode for text analysis. Parsing the JSON response ensures that you can process and utilize the extracted text. Training a custom model is unnecessary for text extraction tasks as the OCR capability is pre-built in Azure AI Vision.
- A. Correct.
Correct: You need an Azure Computer Vision resource to access the OCR capabilities. The endpoint and key are required to authenticate API calls.
- B. Correct.
Correct: The OCR API is the primary tool in Azure AI Vision for extracting text from images.
- C. Incorrect.
Incorrect: Custom Vision is used for object detection and classification, not for text extraction. The OCR API does not require custom model training.
- D. Correct.
Correct: The OCR API returns a JSON response containing the extracted text, which needs to be parsed to retrieve the desired information.
- E. Correct.
Correct: The 'Read' API mode is designed for advanced text extraction, including processing handwritten and printed text.