AI-102 Question 129
Select 2You are tasked with building a solution that processes handwritten notes from scanned documents and extracts the text for further analysis. You decide to use Azure AI Vision for this purpose. Which steps should you include in your solution to accurately convert handwritten text into machine-readable text?
- A
Use the Azure Computer Vision Read API to analyze the handwritten text.
- B
Preprocess the images by enhancing contrast and removing noise before sending them to Azure AI Vision.
- C
Train a custom OCR model in Azure Machine Learning to handle handwritten text.
- D
Validate the extracted text using human review workflows in Azure Cognitive Services.
- E
Use the Azure Form Recognizer API instead of the Read API to extract handwritten text.
Show answer and explanation
Correct answers: A, B
Explanation
To extract handwritten text effectively, you should use the Azure Computer Vision Read API, which is optimized for text extraction, including handwriting. Preprocessing the images by enhancing contrast and removing noise ensures the input quality is high, resulting in better OCR performance. Custom model training or using APIs designed for structured data (like Form Recognizer) is unnecessary, and human review workflows are optional based on the scenario.
- A. Correct.
The Azure Computer Vision Read API is specifically designed to read and extract text from images, including handwritten text, making it a suitable choice for this scenario.
- B. Correct.
Preprocessing the images, such as enhancing contrast and removing noise, improves the accuracy of the OCR process by providing clearer input.
- C. Incorrect.
Training a custom OCR model in Azure Machine Learning is unnecessary in this case because the Azure Computer Vision Read API already supports handwriting recognition.
- D. Incorrect.
While human review workflows can be used for validation in certain scenarios, they are not a required step for basic handwriting recognition using Azure AI Vision.
- E. Incorrect.
The Azure Form Recognizer API is more suited for extracting structured data from forms and tables, not general handwritten text extraction.