Microsoft Azure AI Fundamentals (AI-900)Describe features of computer vision workloads on AzureHard
A city planning department wants to monitor urban development and vegetation growth over time using satellite imagery. They need to generate a precise pixel-level mask for different land cover types (e.g., buildings, roads, trees, water) to accurately calculate area changes. Which computer vision capability is required for this task?
- AInstance Segmentation
- BSemantic Segmentation
- CImage Classification
- DObject Detection
Show answer & explanationAnswer & explanation
Correct answer: B. Semantic Segmentation
Semantic Segmentation classifies each pixel in an image into a predefined class, creating a pixel-level mask for different categories like 'trees' or 'buildings'. This is precisely what is needed to calculate area changes of land cover types.
Why the other options are wrong
- A. Instance Segmentation identifies individual instances of objects and provides pixel-level masks for each, but for land cover types (like 'trees' or 'water'), the goal is typically to segment the general class rather than individual trees or water bodies, making Semantic Segmentation more direct for 'different land cover types'.
- C. Image Classification assigns a single label to the entire image, not pixel-level detail.
- D. Object Detection draws bounding boxes around objects, but doesn't provide pixel-level masks for irregular shapes or overlapping instances of the same class.
Semantic Segmentation
A computer vision technique that assigns a class label to every pixel in an image, effectively segmenting the image into regions corresponding to different object categories.
- Provides pixel-level understanding of an image.
- Useful for precise area calculation, medical imaging, autonomous driving.
- Distinguishes between different classes of objects, but not individual instances within the same class.
Memory trick: Segmentation carves images into meaningful pieces.