INT345:COMPUTER VISION
L:2 T:0 P:2 Credits:3
Course Outcomes: Through this course students should be able to
CO1 :: articulate fundamentals of image processing,image enhancement and filtering.
CO2 :: elaborate various camera models such as pinhole camera model, camera calibration and
lens distortion correction..
CO3 :: derive the concept of epipolar geometry including the epipolar constraint and fundamental
matrix.
CO4 :: apply appropriate feature descriptors to find relevant features from images.
CO5 :: apply various processing techniques on different image categories.
CO6 :: analyze various machine learning algorithms for image classification and clustering.
Unit I
Fundamentals of Image Processing : image acquisition and representation, image file formats and
color models, Overview of computer vision and its applications
Image Enhancement : Histogram equalization, contrast enhancement, Thresholding, The EM
Algorithm, frequency domain techniques
Image Filtering : spatial filtering, fourier transforms, geometric transformations
Unit II
2-D Projective Geometry : planar geometry, the 2D projective plane, projective transformations,
homography properties
Camera Geometry : pinhole cameras, cameras with lenses, CCD cameras, general projective
cameras, affine cameras
Unit III
Stereo Geometry : epipolar geometry, fundamental matrix, motion of camera, normalized 8-point
algorithm, algebric minimization algorithm, geometric distance, experimental evaluation of the
algorithms, linear triangulation methods, geometric error cost function
Unit IV
Feature Detection and Description : feature detection, Harris operator, feature matching, scale
invariant feature transform, speeded up robust features, hessian operator, FAST, BRIEF, ORB, HOG,
texture descriptor, image descriptor, applications of descriptors
Feature Matching and model Fitting : weighted distance function, similarity measures, EMD,
efficient computation of feature matching, K-D tree, LSH, model fitting techniques
Unit V
Color Processing : The physics of color, Human Color Perception, representing color, model for
image color, surface color from image color
Range Image Processing : active range sensors, range data segmentation, range image
registration and model acquistion, object recognition
Unit VI
Clustering and Classification : overview of classification, probabilistic approach, distance based,
discriminant analysis, ANN, clustering techniques, llyod algorithm, k-means clustering
Dimensionality Reduction and Sparse Representation : PCA, LDA, sparse representation, OMP
algorithm, BP algorithm
List of Practicals / Experiments:
Practicals
• Implementing basic image processing operations, such as image filtering, edge detection, and image
enhancement..
• Performing geometric transformations on images, including rotation, scaling, and translation.
• Implementing feature detection algorithms like Harris corner detection or SIFT (Scale-Invariant
Feature Transform) and matching features between images.
Session 2024-25 Page:1/2
• Calibrating a camera using chessboard patterns or other calibration patterns to estimate the intrinsic
and extrinsic camera parameters.
• Implementing stereo matching algorithms for estimating depth from stereo image pairs.
• Estimating optical flow between consecutive video frames to understand motion in videos.
• Implementing image segmentation algorithms, such as thresholding, region growing.
• Building object detection systems using techniques like Haar cascades.
Text Books:
1. COMPUTER VISION: A MODERN APPROACH by DAVID A. FORSYTH, Pearson Education
India
References:
1. PROGRAMMING COMPUTER VISION WITH PYTHON: TOOLS AND ALGORITHMS FOR
ANALYSING IMAGES by JAN ERIK SOLEM, O'REILLY
Session 2024-25 Page:2/2