Lecture-1 - CAP 6411 Computer Vision Systems • Instructor...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: CAP 6411 Computer Vision Systems • Instructor: Dr. Mubarak Shah, [email protected], 238 CSB, http://www.cs.ucf.edu/courses/cap6411 • Office Hours: – 2PM to 3PM Mon, 4PM-5PM Tu, 5:15PM-6:15 PM Wed • Grading – Mid term 20%, Final 25%, Programs 45% , Homework 10% • Recommended Book, but not required. – Digital Video Processing, A. M. Tekalp, Prentice Hall. Multimedia • • • • • Text Graphics Audio Images Video 1 Imaging Configurations • • • • Stationary camera stationary objects Stationary camera moving objects Moving camera stationary objects Moving camera moving objects Video • • • • sequence of images clip mosaic key frames 2 Sequence of Images Clip 3 Mosaic Key Frames 4 Steps in Video Computing • • • • • • • • Acquire Process Analyze Transmit Store Retrieve Browse Visualize (CCD arrays/synthesize (graphics)) (image processing) (computer vision) (compression/networking) (compression/databases) (computer vision/databases) (computer vision/databases) (graphics) Computer Vision: Motion • • • • Motion Detection Motion Measurement (optical flow) Tracking Structure from motion (derive 3-D motion & shape) • Motion Recognition • Motion-based Recognition 5 A Video Clip Consecutive Frame Difference 6 Background Difference Optical Flow Measurement of motion at each pixel 7 Synthetic Images (Random dot stereogram) Results 10 iterations One iteration l=4 Horn-Schunck 8 Image from Hamburg Taxi seq optical flow 9 Lucas-Kanade with Pyramids Optical Flow 10 Video Compression 11 Example Sprite 12 Tracking Tracking & Object Detection In Single Camera PETS-2001 13 Motion Recognition Activities 14 Detecting Violence Video Registration 15 IRS-1C - Washington, DC SPOT - Washington, DC 16 SPOT/IRS-1C Uncorrected SPOT IRS-1C Uncorrected SPOT/IRS-1C Uncorrected 966 m SPOT IRS-1C Uncorrected 17 SPOT/IRS-1C Uncorrected 966 m SPOT IRS-1C Uncorrected IRS-1C/SPOT Registered IRS-1C SPOT 5m Registered 18 Registered IRS-1C to SPOT IRS-1C SPOT Registered Video Segmentation 19 20 Image Processing • Filtering • Compression – – – – MPEG-1 MPEG-2 MPEG-4 MPEG-7 (Multimedia Content Description Interface) 21 Databases • • • • Storage Retrieval Video on demand Browsing – skim – abstract – key frames – mosaics Networking • Transmission • ATM 22 Computer Graphics • Visualization • Image-based Rendering and Modeling • Augmented Reality Contents 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Introduction Motion models Computing 2-D motion Computing 3-D motion and depth Video tracking Video segmentation Video mosaics Visual gesture recognition Visual lipreading Facial expression recognition Human activity recognition Video compression Tools 23 Contents 1. 2. Introduction Motion models 1. 2. 3D Rigid Motion Model Image Projection Models 1. 2. 3. 4. 3. Perspective Projection Orthographic Projection Image Motion Models Relationship between Image Motion 3D Rigid Motion Computing Image motion 1. Computing Local Motion 1. 2. 2. Horn and Schunck Optical Flow Method Lucas and Kanade Computing Global Motion 1. 2. 3. 3. Anandan Szeliski Mann and Piccard Correlation 1. 1. Detailed Contents Computing 3-D motion and depth 1. 2. 2. Motion Compensation Tomassi and Kanade Heeger and Jepson Video tracking 1. Change Detection 1. 2. 3. 4. 5. 6. 7. 8. Generating Background Model Pfinder Mixture of Gaussian W4 Kanade Color based tracking Skin Detection Correlation 24 Detailed Contents 1. Generating Trajectories 1. Token Detection 1. 2. 2. 3. Moaravec’s Interest Operator Lucaas Kanade Operator Motion Correspondence Kalman filter for motion correspondence Detailed Contents 1. Video mosaics 1. 2. Introduction Projection surfaces 1. 2. 3. 3. 4. 2. Plane Sphere Cylinder Static Mosaic Synopsis Mosaic Visual gesture recognition 1. 2. 3. 3. Jim Andrew Pentland Visual lipreading 1. Shawn 25 Detailed Contents 1. Facial expression recognition 1. 2. Yacoob and Black Human activity recognition 1. 2. 3. 3. Polana and Nelson Jim Model-based Activity Recognition Video compression 1. 2. 3. 4. Brief Introduction to JPEG Brief Introduction to MPEG-1 Brief Introduction to MPEG-2 Model-based Compression 1. 2. 3. 5. 6. Video phones Szeliski Making faces MPEG-4 MPEG-7 Detailed Contents 1. Tools 1. 2. 3. 4. 5. 6. 7. 8. Pyramids Interpolation Color Science Image Warping Kalman filter Hidden Markov Models Finite State Machine Dynamic Time Warping 26 ...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online