DAMSL-272 Computer Vision

Syllabus

• Introduction to Computer Vision

• Overview of Image Acquisition and Processing Topics(Sampling, Quantization, Color Perception, Smoothing Filters, Derivatives)
• Overview of Image Analysis Topics(Edge Detection, Segmentation)
• Representation, Analysis, and Synthesis of Textures
• Interest Point Detection(Harris Corner Detector)
• Blob Detection
• Descriptors of Interest Points(SIFT)
• Hough Transform
• Methods for Estimating Parametric Models (Least Squares Method)
• Robust Parameter Estimation Methods (LMedS, RANSAC)
• Alignment of Model-Image Based on Features
• Camera and Lens Models, Projective Geometry
• Camera Calibration
• Epipolar Geometry
• Stereo Vision: The Correspondence Problem and 3D Reconstruction
• Volumetric 3D Reconstruction from Multiple Cameras
• Estimation of 2D Motion(Vertical Optical Flow, Optical Flow)
• Modeling 3D Motion(Motion Field, Eigenmotion)
• Tracking of Linear Dynamic Models
• Tracking with Particle Filters
• Object Detection(Human Body, Face)
• Object Detection
• Object classification

Learning Outcomes

• Having attended and succeeded in the course, the student is able to describe how specific, selected problems in computer vision are addressed in the relevant literature.
• Having attended and succeeded in the course, the student has achieved an in-depth understanding of the mechanisms for solving specific computer vision problems and can explain the reasons that make these mechanisms sufficient for solving other problems.
• Having attended and succeeded in the course, the student can reuse existing methodologies and tools to generate other solutions for solving specific instances of computer vision problems or developing applications.
• Having attended and succeeded in the course, the student can critically evaluate specific problems and perceive them as compositions of a series of individual sub-problems.
•  Having attended and succeeded in the course, the student can combine individual tools and methodologies to achieve the solution of complex computer vision problems.
• Having attended and succeeded in the course, the student can measure/quantitatively assess the quality of solutions to computer vision problems and compare these solutions to other existing ones