⚡FREE Live Master Session: Code Your AI Companion for Kids

    Register for Free →
    Subject 417 • Unit 5 • 4 Marks Theory + 20 Practical Hours

    Unit 5: Computer Vision (CV)Image Processing, Convolution Kernels & CNN Architectures

    The definitive study module for CBSE Class 10 Artificial Intelligence Unit 5. Master digital image representation, 8-bit grayscale vs 24-bit RGB pixel grids, convolution sliding filters, Convolutional Neural Networks (Conv, ReLU, Max Pooling, FC), the Coral Bleaching SDG 14 project, and official handbook MCQs with complete answer keys.

    Practice MCQs
    Board Exam Practice • 12 Questions

    Official Handbook MCQs & Answer Key

    Click any option to instantly see if you're correct with a detailed explanation.

    1

    What is the primary difference between Computer Vision (CV) and Image Processing (IP)?

    2

    Which of the following is an example of Image Classification with Localization?

    3

    What is a 'Pixel' in digital image processing?

    4

    In an 8-bit Grayscale image, what do the pixel values 0 and 255 represent respectively?

    5

    An RGB image consists of how many color channels and what is the total theoretical color depth in standard 24-bit representation?

    6

    What mathematical operation is performed when a Convolutional Kernel (filter) moves across an image?

    7

    Why are standard Multi-Layer Perceptrons (MLPs / Dense Networks) unsuitable for complex computer vision tasks compared to CNNs?

    8

    What is the primary function of the ReLU (Rectified Linear Unit) activation layer in a CNN?

    9

    Which layer in a CNN reduces the spatial dimensions (height and width) of feature maps while retaining dominant features?

    10

    In the Coral Bleaching AI project highlighted in the CBSE handbook, which UN Sustainable Development Goal (SDG) is directly supported?

    11

    What is the purpose of the 'Flatten' layer positioned between the convolutional/pooling layers and the dense layers in a CNN?

    12

    Which no-code web platform developed by Google is recommended in the CBSE curriculum for training beginner Computer Vision models using a webcam?

    Fast Revision Summary

    Unit 5 Quick Recall Cheat Sheet

    IP vs CV: Image in → Image out is Image Processing. Image in → Label / Semantic Insight is Computer Vision.
    Grayscale vs RGB: Grayscale = 1 channel (0–255). RGB = 3 channels ($256^3$ = ~16.7 million colors).
    Convolution Kernel: Sliding filter performing dot product multiplication and summation to extract feature maps.
    5 CNN Layers: Conv (Feature Extraction) → ReLU (Non-linearity) → Pooling (Downsampling) → Flatten (1D Vector) → Dense + Softmax (Classification).
    Max Pooling: Extracts maximum pixel value in each window, shrinking dimensions and providing translation invariance.
    Coral Bleaching AI: Classification of marine coral health (Healthy vs Bleached vs Dead) supporting SDG 14: Life Below Water.
    Clear Your Doubts

    Frequently Asked Questions (FAQ)

    Ace Your Board Exams • Subject 417

    Master Computer Vision & AI with 1:1 Live Mentorship

    Get 1-on-1 personalized guidance from certified AI educators. Master CNN models in Python, OpenCV practicals, Google Teachable Machine projects, and complete scoring blueprints for your CBSE Class 10 Board Exams.