MR4015 Advanced Computer Vision Syllabus:

MR4015 Advanced Computer Vision Syllabus – Anna University PG Syllabus Regulation 2021

COURSE OBJECTIVES

1. To understand the various fundamental mathematics behind computer vision algorithms.
2. To expose students to various image formation and camera calibration techniques
3. To expose students to various 3D surface reconstruction algorithms.
4. To impart knowledge on stereo vision and structure from motion.
5. To impart knowledge on applying the computer vision techniques to robots

UNIT – I BASIC CONCEPTS FOR COMPUTER VISION

Sampling Theorem – Numerical Differentiation – Differential Geometry – Singular Value Decomposition – Robust Estimators and Model Fitting

UNIT – II IMAGE FORMATION AND CAMERA CALIBRATION

Projective Geometry – Imaging through lenses and pin-hole – Basic Photometry – Basic model of imaging geometry – Ideal Camera – Camera with intrinsic parameters – Approximate camera models – Camera Calibration – Methods and Procedure

UNIT – III SURFACE RECONSTRUCTION TECHNIQUES

Depth Perception in Humans, Cues – Shape from Texture, Shading, Focus, Defocus, Structured Light Reconstruction – Time of Flight Methods

UNIT – IV COMPUTATIONAL STEREO AND MOTION

Computational Stereopsis – Geometry, parameters –correlation-based methods, feature based methods – Epipolar Geometry, eight-point algorithm – Reconstruction by triangulation, scale factor and up to a projective transformation – Visual Motion – Motion field of rigid objects – Optical Flow – Estimation of motion field – 3D structure and motion from sparse and dense motion fields – Motion based segmentation.

UNIT – V ROBOT VISION

Visual Tracking – Kalman Filtering and Sequential Monte Carlo – Visual SLAM, solutions, EKFSLAM, Fast SLAM – 3D SLAM – Advanced Visual Servoing, hybrid visual servo, partitioned visual servo.

TOTAL: 45 PERIODS

COURSE OUTCOMES:

Upon completion of this course, the students will be able to:
CO1: Understand the basic concepts behind computer vision algorithms.
CO2: Understand various image formation and camera calibration techniques.
CO3: Understand various 3D surface reconstruction algorithms.
CO4: Understand stereo vision and structure from motion.
CO5: Apply the computer vision techniques to robots

REFERENCES:

1. Eugene Hecht, A.R. Ganesan “Optics”, Fourth Edition, 2001.
2. Emanuele Trucco, Alessandro Verri, “Introductory Techniques For 3D Computer Vision”, First Edition, 1998.
3. Boguslaw Cyganek, J. Paul Siebert, An Introduction To 3D Computer Vision Techniques and Algorithms, First Edition, 2009.
4. Yi Ma, Jana Kosecka, Stefano Soatto, Shankar Sastry, An Invitation to 3-D Vision from Images to Models, First Edition, 2004.