docIntroduction - Bilkent University Computer Engineering Department
download 
Şikayet Transkript
Introduction - Bilkent University Computer Engineering Department
Introduction Selim Aksoy Department of Computer Engineering Bilkent University [email protected] What is computer vision? Analysis of digital images by a computer. Stockman and Shapiro: making useful decisions about real Trucco and Verri: computing properties of the 3D world physical objects and scenes based on sensed images. from one or more digital images. Ballard and Brown: construction of explicit, meaningful description of physical objects from images. Forsyth and Ponce: extracting descriptions of the world from pictures or sequences of pictures. CS 484, Spring 2007 ©2007, Selim Aksoy 2 Why study computer vision? Possibility of building intelligent machines is fascinating. Capability of understanding the visual world is a prerequisite for such machines. Much of the human brain is dedicated to vision. Humans solve many visual problems effortlessly, yet we have little understanding of visual cognition. CS 484, Spring 2007 ©2007, Selim Aksoy 3 Why study computer vision? An image is worth 1000 words. Images and videos are everywhere. Fast growing collections and many useful applications. Goals of vision research: Give machines the ability to understand scenes. Aid understanding and modeling of human vision. Automate visual operations. CS 484, Spring 2007 ©2007, Selim Aksoy 4 Challenge What do you see in the picture? A hand holding a man A hand holding a shiny sphere An Escher drawing Adapted from Octavia Camps, Penn State CS 484, Spring 2007 ©2007, Selim Aksoy 5 Perception and grouping Subjective contours CS 484, Spring 2007 ©2007, Selim Aksoy 6 Perception and grouping Subjective contours Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 7 Perception and grouping Adapted from Gonzales and Woods CS 484, Spring 2007 ©2007, Selim Aksoy 8 Perception and grouping Adapted from Gonzales and Woods CS 484, Spring 2007 ©2007, Selim Aksoy 9 Perception and grouping Occlusion Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 11 Perception and grouping The shape of junctions constrains the possible interpretations of the scene. Ambiguous: paint and surface boundaries can be confused. Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 12 Recognition How can different cues such as color, texture, shape, motion, etc., can be used for recognition? Which parts of image should be recognized together? How can objects be recognized without focusing on detail? How can objects with many free parameters be recognized? How do we structure very large model bases? CS 484, Spring 2007 ©2007, Selim Aksoy 13 Color Adapted from Martial Hebert, CMU CS 484, Spring 2007 ©2007, Selim Aksoy 14 Texture CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from David Forsyth, UC Berkeley 15 Segmentation Original Images Color Regions Texture Regions Line Clusters Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 16 Segmentation CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from Jianbo Shi, U Penn 17 Shape Recognized objects Adapted from Enis Cetin, Bilkent University Model database CS 484, Spring 2007 ©2007, Selim Aksoy 18 Motion Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 19 Recognition Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 20 Recognition Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 21 Recognition Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 22 Recognition Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 23 Recognition Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 24 Recognition CS 484, Spring 2007 ©2007, Selim Aksoy 25 Recognition Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 26 Detection Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 27 Detection Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 28 Detection Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 29 Parts and relations Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 30 Parts and relations Adapted from Michael Black, Brown University CS 484, Spring 2007 ©2007, Selim Aksoy 31 Context Adapted from Antonio Torralba, MIT CS 484, Spring 2007 ©2007, Selim Aksoy 32 Context Adapted from Antonio Torralba, MIT CS 484, Spring 2007 ©2007, Selim Aksoy 33 Context Adapted from Derek Hoiem, CMU CS 484, Spring 2007 ©2007, Selim Aksoy 34 Context Adapted from Derek Hoiem, CMU CS 484, Spring 2007 ©2007, Selim Aksoy 35 Stages of computer vision Low-level image Æ image Mid-level image Æ features / attributes Image analysis / image understanding High-level features Æ “making sense”, recognition CS 484, Spring 2007 ©2007, Selim Aksoy 36 Low-level sharpening blurring Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 37 Low-level Canny original image Mid-level edge image ORT data structure edge image CS 484, Spring 2007 circular arcs and line segments ©2007, Selim Aksoy Adapted from Linda Shapiro, U of Washington 38 Mid-level K-means clustering (followed by connected component analysis) regions of homogeneous color original color image data structure Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 39 Low-level to high-level low-level edge image mid-level high-level consistent line clusters Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 40 Applications Industrial inspection, quality control Medical image analysis Remote sensing Security Surveillance Biometrics Target recognition Tracking CS 484, Spring 2007 Robotics Document analysis Multimedia Assisted living Human-computer interfaces ©2007, Selim Aksoy 41 Medical image analysis CT image of abdomen kidney kidney Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 42 Medical image analysis Cancer detection and grading CS 484, Spring 2007 ©2007, Selim Aksoy 43 Medical image analysis Slice of lung CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from Linda Shapiro, U of Washington 44 Land cover analysis CS 484, Spring 2007 ©2007, Selim Aksoy 45 Object recognition Recognition of buildings and building groups CS 484, Spring 2007 ©2007, Selim Aksoy 46 Content-based retrieval Finding similar regions: airports CS 484, Spring 2007 ©2007, Selim Aksoy 47 Tracking Adapted from Octavia Camps, Penn State CS 484, Spring 2007 ©2007, Selim Aksoy 48 Tracking Adapted from Martial Hebert, CMU CS 484, Spring 2007 ©2007, Selim Aksoy 49 Tracking Adapted from Martial Hebert, CMU CS 484, Spring 2007 ©2007, Selim Aksoy 50 Biometrics Adapted from Anil Jain, Michigan State CS 484, Spring 2007 ©2007, Selim Aksoy 51 Biometrics Adapted from Anil Jain, Michigan State CS 484, Spring 2007 ©2007, Selim Aksoy 52 Robotics Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 53 Robotics Adapted from Steven Seitz, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 54 Document analysis Digit recognition, AT&T labs http://www.research.att.com/~yann Adapted from Steven Seitz, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 55 Document analysis Adapted from Shapiro and Stockman CS 484, Spring 2007 ©2007, Selim Aksoy 56 Document analysis CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from Linda Shapiro, U of Washington 57 Scene classification CS 484, Spring 2007 ©2007, Selim Aksoy 58 Organizing image archives Adapted from Pinar Duygulu, Bilkent University CS 484, Spring 2007 ©2007, Selim Aksoy 59 Content-based retrieval CS 484, Spring 2007 ©2007, Selim Aksoy 60 Content-based retrieval CS 484, Spring 2007 ©2007, Selim Aksoy 61 Content-based retrieval CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from David Forsyth, UC Berkeley 62 Face detection and recognition CS 484, Spring 2007 ©2007, Selim Aksoy 63 Object recognition CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from Rob Fergus, MIT 64 3D scanning Adapted from Linda Shapiro, U of Washington CS 484, Spring 2007 ©2007, Selim Aksoy 65 3D reconstruction Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 66 3D reconstruction Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 67 Motion capture CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from Linda Shapiro, U of Washington 68 Visual effects CS 484, Spring 2007 ©2007, Selim Aksoy Adapted from Linda Shapiro, U of Washington 69 Mozaic Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 70 Mozaic Adapted from David Forsyth, UC Berkeley CS 484, Spring 2007 ©2007, Selim Aksoy 71 Critical issues What information should be extracted? How can it be extracted? How should it be represented? How can it be used to aid analysis and understanding? CS 484, Spring 2007 ©2007, Selim Aksoy 72
Benzer belgeler
Introduction - Bilkent University Computer Engineering Department
Capability of understanding the visual world is a
prerequisite for such machines.
Much of the human brain is dedicated to vision.
Humans solve many visual problems effortlessly,
yet we have little ...
cs484_intro [Compatibility Mode] - Bilkent University Computer
![cs484_intro [Compatibility Mode] - Bilkent University Computer](http://s1.doczz.biz.tr/store/data/000057576_1-26d821749e886de04e27a5422d8297db-70x70.png "cs484_intro [Compatibility Mode] - Bilkent University Computer")
Adapted from L. Fei-Fei, R. Fergus, A. Torralba
Brockers, Roland........................................................ 262
Linear Filtering – Part II - Bilkent University Computer Engineering
Correlation can also be used for matching.
If we want to determine whether an image f
contains a particular object, we let h be that
object (also called a template) and compute the
correlation betw...