A parking garage operator wants a camera system that, for each captured frame, draws a separate bounding box around every vehicle in view and labels each one by type (car, truck, motorcycle), so the system can report how many of each type are currently parked. Which Azure AI Vision capability should the team use?
Select an answer to reveal the explanation.
Short Explanation
Think about what the garage actually needs: not just a single word describing the whole picture, and not a text readout, but a separate box drawn around each vehicle with its own label so they can be counted individually. A tool that only stamps one label on the whole photo can tell you 'yes, there are vehicles here' but can't tell you how many trucks versus cars are present, because it never breaks the image into individual items. A text-reading tool is aimed at signs and plates, not at recognizing vehicle shapes. And a tool that paints every vehicle pixel the same color is useful for knowing where the 'vehicle region' is, but it blurs neighboring cars together into one blob instead of giving you separate, countable instances. What the garage needs is a capability built specifically to find and label each object on its own, which is why it fits this job better than the others.
Full Explanation
The correct answer is A. Object detection returns a bounding box and a label for each individual item in an image, which is exactly what the garage needs to isolate and count multiple vehicles of different types within a single frame. Option B is incorrect because image classification assigns one label to the entire image, such as 'parking lot,' and cannot localize or distinguish multiple vehicles within that same frame. Option C is incorrect because OCR extracts printed or handwritten text, such as license plate numbers or signage, and has no mechanism for identifying or counting vehicle types. Option D is incorrect because semantic segmentation classifies every pixel into a category, such as marking all 'vehicle' pixels as one class, but it does not separate touching or overlapping vehicles into distinct, countable instances the way object detection's individual bounding boxes do. Because the garage's requirement is discrete counts per labeled vehicle, only object detection satisfies both the localization and the per-instance labeling the scenario demands.