COMP 9517 Computer Vision

1. COMP 9517 Computer Vision Introduction COMP 9517 S2, 2009

2. Computer Vision • Computer vision has been around since 1960’s, but it only recently became possible to build useful vision system. • Computer and imaging system have become much cheaper. • Our understanding of basic geometry and physics underlying vision has been improved significantly. • It is a great time to be studying computer vision. COMP 9517 S2, 2009

3. What is Computer Vision? • What are the goals of computer vision? • What are the applications? • What are computer vision processes? COMP 9517 S2, 2009

4. Goals of Computer Vision • Make useful decisions about real physical objects and scenes based on sensed images or sequences of images. • Alternative: goal is the construction of scene descriptions from images. • Use statistical methods to process data using models constructed with the aid of geometry, physics and learning theory. COMP 9517 S2, 2009

5. Computer Vision Topics • Requires a solid understanding of camera and of the physical process of image, • to obtain simple inferences from individual pixel values, • to combine the information available in multiple images into a coherent whole, • to enforce some order on groups of pixels to separate them from each other or infer shape information, • to recognise objects using geometric information or probabilistic techniques. COMP 9517 S2, 2009

6. Critical Issues • Sensing: how do sensors obtain images of the world? • Encoded Information: how do image yield information of the scene, such as color, texture, shape, motion, etc.? • Representations: what representations are appropriate to describe objects? • Algorithms: what algorithms process image information and construct scene descriptions? COMP 9517 S2, 2009

7. CV Applications • Vision-based HCI • EyeMouse: a vision-based eye control system • To use human head and eyes to control computers, so how? • Computer vision and an webcam to track the eyes and head • Shakes and winks to control a mouse pointer on the screen • Face expression recognition • Challenge: clutter and real time • Game Controller: Cam-Trax COMP 9517 S2, 2009

8. CV Applications • Geographical: GIS • Interpreting satellite images • Road detection for creating maps • Edge detection, Road edge classification and linking • Challenge: complex and wide scene, occlusion, low resolution or large data size. COMP 9517 S2, 2009

9. CV Applications • Medical Imaging • Enhance imagery, or identify important phenomena or events, or visualise information obtained by imaging • Parenchymal bands: linear structures touching the lung boundary • Segment and classify candidate regions into positive (parenchymal bands) and negative (others) class • Challenge: • Often attached to other structures, in this case a nodular mass • Similar appearance to blood vessels COMP 9517 S2, 2009

10. CV Applications • Video Surveillance • Traffic Monitoring • Object tracking • Action recognition, driving, stopping, etc • Vehicle speed • Counting • Challenge: occlusion, illumination changes and non-linear speed COMP 9517 S2, 2009

11. CV Applications • Image/video retrieval • Content-based retrieval • Search engine • Challenge: big data volume, semantic COMP 9517 S2, 2009

12. CV Applications • Text Recognition • Converting information from paper documents into digital form • Challenge: semantic interpretation I looked as hard as I could see, beyond 100 plus infinity an object of bright intensity- it was the back of me! COMP 9517 S2, 2009

13. Computer Vision Processes • Low level processes • use little knowledge of image content • include image compression, noise filtering, edge extraction, ... • use data which resemble the input image, eg. matrix of picture elements • High level processes • based on knowledge, goals, plans • use Artificial Intelligence methods • simulate human cognition and decision making based on information in the image • cognitive processes, geometric models, goals, plans,... COMP 9517 S2, 2009

14. Low Level Vision • almost entirely digital image processing • sensing: image capture and digitisation • pre-processing: improve image quality: suppress noise, enhance object features, edge extraction • image segmentation: separate objects from background, partition image into objects of interest • description: compute features which differentiate objects- also called feature extraction • Classification: assign labels to image segments (regions) COMP 9517 S2, 2009

15. High Level Vision • About knowledge construction, representation and inference • recognition: identification of objects • interpretation: assign meaning to groups of recognized objects • scene analysis COMP 9517 S2, 2009