DBATU · Final Year B.Tech Artificial Intelligence & Machine Learning · Non-NEP
Advanced Computer Vision
BTAIC702 · Semester VII · 4 credits · Continuous assessment 20 + mid-semester 20 + end-semester 60
Unit 1: Introduction to Computer Vision and Image Processing
7 hoursComputer vision and its applications; image representation and processing; filtering and convolution; contrast stretching, histogram specification, adaptive histogram equalisation and wavelet enhancement; linear, non-linear and Laplacian filters.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Computer vision pipeline
- Digital image representation
- Histogram equalisation
- Wavelet-based image enhancement
- Convolution and smoothing filters
- Sharpening filters
Unit 2: Image Transformation and Restoration
7 hoursGeometric transformations; DFT, DCT, DST, Walsh–Hadamard, slant and Haar transforms; noise models and restoration with mean, median and midpoint filters.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Image transforms: DFT, DCT, Walsh-Hadamard and Haar
- Noise models and noise removal
- Restoration with mean and order-statistic filters
Unit 3: Segmentation, Texture and Motion Analysis
8 hoursEdge detection (Prewitt, Sobel, Canny), thresholding and region-based segmentation; texture analysis with GLCM, LBP, Gabor filters and Laws' measures; optical flow by Lucas–Kanade and Horn–Schunck, FlowNet, background subtraction and motion segmentation.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Gradient edge detectors
- Canny edge detector
- Thresholding and Otsu's method
- Region-based segmentation
- Texture features: GLCM, LBP and Gabor filters
- Optical flow
- Background subtraction and motion segmentation
Unit 4: Feature Matching Algorithms
7 hoursSIFT, SURF and BRISK features; feature datasets, bag of words and vector quantisation; classifying features with SVM, KNN and random forests.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- SIFT, SURF and BRISK features
- Bag of visual words
- Classifying image features: SVM, KNN and random forests
Unit 5: Computer Vision with Deep Learning
7 hoursCNN image classification, attention and vision transformers; GANs; object detection with R-CNN, Fast and Faster R-CNN, Mask R-CNN, SSD and YOLO; U-Net semantic segmentation; centroid-based tracking.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- CNNs for image classification
- Vision transformers
- GANs for images
- Object detection
- R-CNN family, SSD and YOLO
- Semantic segmentation with U-Net
- Centroid-based object tracking
Unit contents are summarised from the official DBATU syllabus . AiCanCode.org is not affiliated with Dr. Babasaheb Ambedkar Technological University (DBATU).