CPE 631 (2009) / CPE 488 Machine Vision Computer Engineering Department King Mongkut's University of Technology, Thonburi Course Information Instructor: Website: Time: Office Hours: TA.: Office Hours: Suthep Madarasmi, Ph.D. (suthep@kmutt.ac.th) www.cpe.kmutt.ac.th/~suthep/cpe631 Friday 17:00-19:00, CB 40805 Friday 16:00 – 17:00 Ms. Varin Chouvatut (varin@cpe.kmutt.ac.th) Thursdays 13:30 – 14:30 at Sigma Lab. Course Overview To introduce students to the concepts of machine vision touching on areas of computer graphics, image processing, artificial intelligence, biological vision, neural networks, pattern recognition and robot vision. The course will be project-oriented consisting of a lecture to introduce the subject matter followed by discussions on computer vision applications and assignments to learn the subject matter. Computer vision can be viewed as the inverse problem of computer graphics: the objective of computer graphics is to generate images using a model of the world whereas the objective of computer vision is to arrive at a description of the world using images. Image processing will be covered extensively with topics such as edge finding, image enhancement, image segmentation, and clustering. Linear, non-linear, and stochastic optimization methods will be introduced for use in solving computer vision problems. The inverse optics problem in computer vision will be discussed including stereo vision, shape from shading, and other Shape from X algorithms. Finally, we will discuss several algorithms for image understanding such as scene interpretation, object recognition, and face recognition. Assignments 1. Simple Thai OCR Competition: Correlation, edge detection, thinning, image segmentation, and template matching. Undergraduates can form groups of 2 persons for the assignment. Graduates must do individually. (15%) 2. Term project will be on a computer vision topic of choice requiring: Proposal: 1-2 pages and 5 minutes presentation. Project: 20 minutes presentation for project work, 30 minutes for literature review lecture. Text Jain R., R. Kasturi, and B.G. Schunck, Machine Vision, McGraw-Hill. Course Topics 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Overview of Computer Vision Course and Assignments Overview Image Formation and Sensing 3-D Computer Graphics and Visual Realism Digital Images: bw, grayscale, and color Binary Image Processing: Low-level Image Filtering and Edge Detection Regions, Image Segmentation, Texture Segmentation Blob Coloring Contours and Boundary Detection General Hough Technique and Applications 3-D Computer Graphics Models Revisited Optics, Curves and Surface Optimization: Pseudo-Inverse, Hough Technique. Optimization: Energy Minimization Gradient Descent, Image Cleanup Back Propagation Neural Networks Computing Optical Flow Optimization: Bayesian Probability, Pixel Lattice Gibbs Sampler and Simulated Annealing Depth from Stereo Vision Optimization: Genetic Algorithms Optimal Material Consumption Depth & Shape from X, Texture, Contour, Stereo Calibrations, Structure from Motion, Object Tracking Shape from Shading Camera Pose Estimation and Augmented Reality Object Recognition Models Grading Programming Assignment 1 Assignment 2 (Term Project) Best 9 of 11 Quiz 15% 20% 65%