Course Details
Subject {L-T-P / C} : EE6148 : Computer Vision { 3-0-0 / 3}
Subject Nature : Theory
Coordinator : Prof. Dipti Patra
Syllabus
Module 1:
IMAGE FORMATION MODELS: Monocular imaging system, Orthographic & Perspective Projection, Camera model and Camera calibration, Binocular imaging systems, Stereo Vision, Epipolar Geometry, Multiple views geometry, Structure determination, shape from shading, Photometric Stereo, Depth from Defocus, Construction of 3D model from images. [8 hours]
Module 2:
IMAGE DESCRIPTION: Boundary descriptors, Shape numbers, Fourier descriptors, Statistical moments, Regional descriptors, Topological descriptors, Texture, Moment invariants, Use of principal components for description, Relational descriptors, SHAPE REPRESENTATION: Contour based representation, Region based representation, De-formable curves and surfaces, Snakes and active contours, Level set representations, Fourier, and wavelet descriptors, Medial representations. [10 hours]
Module 3:
FEATURE DETECTION AND MATCHING: Points and patches, Edges and contours, Contour tracking, Lines and vanishing points, SIFT, SURF IMAGE ALIGNMENT AND STITCHING Pairwise alignment, Image stitching, Global alignment, Compositing, MOTION ESTIMATION Motion models, Parametric motion, Optical flow, Motion tracking. [10 hours]
Module 4:
IMAGE UNDERSTANDING: Pattern recognition methods, Object recognition, Image Classification, Face detection and recognition, 3D shape models of faces. COMPUTER VISION APPLICATIONS: Surveillance, foreground-background separation, human gait analysis, In-vehicle vision system: locating roadway, road markings, identifying road signs, locating pedestrians. [8 hours]
Course Objectives
- To describe the foundation of image formation and understand the geometric relationships between 2D images and the 3D world.
- To understand and implement various methods for image description and representation, alignment, and matching in images.
- To understand and implement the technical approaches for object/scene detection, recognition and categorization from images.
- To have an exposure to advanced concepts, including state of the art machine learning architectures, in all aspects of computer vision.
Course Outcomes
Upon completion of this course, students will be able to: <br /> <br />CO1: Be familiar with both the theoretical and practical aspects of image computing, foundation of image formation, measurement, and analysis. <br /> <br />CO2: Understand the geometric relationships between 2D images and the 3D world. <br /> <br />CO3: Understand and implement the methods for robust image description, representation, image alignment and matching. <br /> <br />CO4: Have gained exposure to object/scene recognition, tracking, classification from images. <br /> <br />CO5: Identify, demonstrate and apply their knowledge by analyzing computer vision problems and employing (or proposing) effective solutions for various applications
Essential Reading
- David A. Forsyth and Jean Ponce,, Computer Vision: A Modern Approach, Pearson Education , 2008
- Richard Szeliski, Computer Vision: Algorithms and Applications, Springer, 2010
Supplementary Reading
- L.G. Shapiro, G. C. Stockman, Computer Vision, Prentice Hall
- T. Morris, Computer Vision and Image Processing, Palgrave McMillan
Journal and Conferences
- IEEE Transaction on Pattern Analysis and Machine Intelligence
- IEEE conference on Computer Vision & Pattern Recognition