Digital Image Processing (DIP)
Data-driven exam preparation guide and searchable question bank based on 11 past papers.
Frequency Analysis
| Topic | Unit | Times Asked | Question Type |
|---|---|---|---|
| Histogram Processing (Equalization/Specification) | 2 | 13 | 8-mark / 15-mark |
| Steps in DIP / Block Diagram | 1 | 12 | 8-mark / 10-mark Theory |
| Smoothing/Sharpening Spatial Filters (Masks) | 2 | 11 | 8-mark / 15-mark |
| Variable Length Coding (Huffman) | 4 | 11 | 8-mark / 15-mark |
| Image Classification using SVM & ANN | 5 | 11 | 8-mark / Theory |
| Object Detection using CNN-RCNN | 5 | 9 | 8-mark / 15-mark |
| Elements of Visual Perception & Color Models | 1 | 9 | 8-mark / Theory |
| Wiener Filtering vs. Inverse Filtering | 3 | 9 | 8-mark / Theory |
| 2D Transforms (DFT, DCT, Haar) | 1 | 8 | 8-mark / Theory |
| Degradation Model & Noise Models | 3 | 8 | 8-mark / Theory |
| Extracting Interest Points (Harris, SIFT, SURF) | 5 | 8 | 8-mark / Theory |
| PCA for Dimensionality Reduction | 5 | 7 | 15-mark / 8-mark |
15-Mark Question Predictions
The Question Bank and recent papers confirm a massive shift towards Applied Algorithms and Dimensionality Reduction.
Prediction 1: Histogram Equalization (Unit 2)
Remains the safest bet. You will likely be given a 6x6 or 8x8 image distribution table and asked to perform equalization step-by-step and plot the graphs.
Prediction 2: PCA Dimensionality Reduction (Unit 5)
Strongly trending in recent papers. Expect a small dataset where you must calculate the covariance matrix, find eigenvalues/eigenvectors, and project data.
Prediction 3: System Design / Application (Unit 5)
Based on the Question Bank, questions like "Design a system for detecting driver drowsiness using image processing techniques" (CNN/Haar Cascade application) are highly probable.
Full Question Bank Database
Type: