AI-900 Question 159
Select 4You are designing a solution for a retail store that needs to automatically identify different types of products (e.g., bottles, boxes, and cans) on a conveyor belt using Azure AI services. Which features of an object detection solution should you consider to meet this requirement?
- A
Ability to detect multiple objects in a single image
- B
Capability to identify the exact position (bounding boxes) of objects within an image
- C
Integration with text recognition to extract product labels
- D
Availability of pre-trained models for common objects
- E
Real-time processing support for live video feeds
Show answer and explanation
Correct answers: A, B, D, E
Explanation
Object detection solutions are designed to identify and locate multiple objects within images or video streams using features like bounding boxes, pre-trained models, and real-time processing. These capabilities are essential for scenarios such as automated product identification in retail. Text recognition, while useful for related tasks, is not a direct feature of object detection.
- A. Correct.
Correct. Object detection solutions must be able to detect multiple objects in a single image, as the conveyor belt may carry various items simultaneously.
- B. Correct.
Correct. Identifying the exact position of objects (using bounding boxes) is a key feature of object detection to pinpoint where each product is located in an image.
- C. Incorrect.
Incorrect. While text recognition is useful for extracting product labels, it is not a core feature of object detection itself. This is a capability of Optical Character Recognition (OCR), which is a separate AI task.
- D. Correct.
Correct. Pre-trained models for common objects can speed up the implementation of an object detection solution by providing ready-to-use capabilities.
- E. Correct.
Correct. Real-time processing is crucial for solutions like monitoring live video feeds on a conveyor belt to detect and act on objects dynamically.