Microsoft Certified Azure AI Fundamentals · Free Practice Question Medium
Question 15
A developer is preprocessing images for a fruit classification model. They convert all images to grayscale to reduce computational complexity.
What is a potential drawback of this approach?
-
A
Grayscale images require larger storage space.
-
B
The model may lose the ability to distinguish fruits with similar shapes but different colors (e.g., lemons vs. bananas).
-
C
Grayscale simplifies edge detection, improving accuracy.
-
D
RGB channels are incompatible with convolutional neural networks (CNNs).
Reveal correct answer
Correct answer: B
Explanation
In a scenario where a developer preprocesses images for a fruit classification model by converting all images to grayscale, the goal is to reduce computational complexity. However, this approach comes with a notable drawback related to the information loss:
Loss of Color Information:
Converting images from RGB to grayscale reduces the image from three color channels (red, green, blue) to a single channel. While this simplifies the data and decreases processing requirements, it removes the color cues that can be critical in distinguishing between fruits. Many fruits may have similar shapes and textures, but their colors are key distinguishing features. For example, two fruits with similar contours—like lemons and bananas—might differ primarily in color. When images are converted to grayscale, the classifier loses access to these color differences, potentially reducing the accuracy of the classification.Impact on Model Performance:
Although grayscale conversion can speed up processing and lower storage needs, the trade-off is that the model might not perform as well when differentiating between classes that rely heavily on color distinctions. In computer vision tasks such as fruit classification, retaining color information is often essential for achieving higher accuracy.
Why This Drawback Matters:
While grayscale images are computationally lighter and can be effective for certain tasks (especially when shape is the primary feature), in applications like fruit classification where color is a critical differentiator, the loss of color information can severely impair the model’s ability to correctly identify and classify different fruits.
Further Reading:
For more on image preprocessing and the role of color in computer vision tasks, review the Computer Vision documentation on Microsoft Learn
Microsoft's Educational Content Around This Topic For The AI-900 Exam:
A.
Grayscale reduces data (1 channel vs. 3), lowering storage needs.
B.
Grayscale removes color information, which is critical for distinguishing objects like yellow bananas vs. green/red apples.
C.
Edge detection can work in grayscale, but color nuances are lost.
D.
CNNs process RGB and grayscale images.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
