CS 591 E / CS 791 CS 591 E / CS 791 L L (CRN: 18390 / (CRN: 18390 / 18490) 18490) Computer Vision Computer Vision Instructor: Guodong Guo Instructor: Guodong Guo [email protected] [email protected]
Jan 19, 2015
CS 591 E / CS 791 CS 591 E / CS 791 LL
(CRN: 18390 / (CRN: 18390 / 18490)18490)
Computer VisionComputer VisionInstructor: Guodong GuoInstructor: Guodong Guo
[email protected]@mail.wvu.edu
Welcome!Welcome!
IntroductionsIntroductions Administrative MattersAdministrative Matters Course OutlineCourse Outline Applications of Computer VisionApplications of Computer Vision Computer Vision FocusComputer Vision Focus Computer Vision PublicationsComputer Vision Publications
Journals Journals ConferencesConferences
InstructorInstructor
Guodong GuoGuodong Guo Ph.D. in CS from UW-MadisonPh.D. in CS from UW-Madison http://www.cs.wisc.edu/~gdguo Major Research InterestMajor Research Interest
Computer Vision, Machine Learning, Computer Vision, Machine Learning, Pattern Recognition, Biometrics, Pattern Recognition, Biometrics, Multimedia, and HCIMultimedia, and HCI
About You …About You …
What do you know already?What do you know already? C/C++ (Visual C++)C/C++ (Visual C++) MatlabMatlab ImagesImages OpenCVOpenCV
http://sourceforge.net/projects/opencvlibrary/
Install OpenCV in your PC or laptop, Install OpenCV in your PC or laptop,
Read the manual introductionRead the manual introduction
Try to load and save images (homework #0)Try to load and save images (homework #0)
OutlineOutline
IntroductionsIntroductions Administrative MattersAdministrative Matters Course OutlineCourse Outline Applications of Computer VisionApplications of Computer Vision Computer Vision FocusComputer Vision Focus Computer Vision PublicationsComputer Vision Publications
Meeting TimesMeeting Times
LecturesLectures M 17:00-19:30 pm M 17:00-19:30 pm Room ESB-E 449Room ESB-E 449
Office hoursOffice hours TR 1:00-2:00 pm (ESB 753)?TR 1:00-2:00 pm (ESB 753)? Or by appointmentOr by appointment
GradingGrading
The final grade depends on:The final grade depends on: Homework and programming Homework and programming
assignments: 40%assignments: 40% Exams (Midterm): 40%Exams (Midterm): 40% Final project (may include class Final project (may include class
presentation): 20%presentation): 20% Class participation: (-5%, if absent >= Class participation: (-5%, if absent >=
3times)3times) Extra: 1~10% (for creative ideas, paper Extra: 1~10% (for creative ideas, paper
submission, etc.)submission, etc.)
TextbookTextbook
Computer Vision: A Computer Vision: A Modern ApproachModern Approach, , 22thth Edition, by Edition, by David Forsyth and David Forsyth and Jean Ponce, Jean Ponce, Prentice Hall, 2003Prentice Hall, 2003
Look at the SyllabusLook at the Syllabus
Course ObjectivesCourse Objectives Expected learning outcomesExpected learning outcomes Detailed list of topics (maybe Detailed list of topics (maybe
updated)updated)
OutlineOutline
IntroductionsIntroductions Administrative MattersAdministrative Matters Course OutlineCourse Outline Applications of Computer VisionApplications of Computer Vision Computer Vision FocusComputer Vision Focus Computer Vision PublicationsComputer Vision Publications
What is Computer What is Computer Vision?Vision?
Given an image or more, extract Given an image or more, extract properties of the 3D world properties of the 3D world
•Traffic sceneTraffic scene• Number of vehiclesNumber of vehicles• Type of vehiclesType of vehicles• Location of closest obstacleLocation of closest obstacle• Assessment of congestionAssessment of congestion
Computer Vision vs. Computer Vision vs. GraphicsGraphics
3D3D2D implies information loss2D implies information loss
sensitivitysensitivity to errors to errors need for need for modelsmodels
graphics
vision
Computer Vision vs. Computer Vision vs. BiometricsBiometrics
BiometricsBiometrics comprises methods for comprises methods for uniquely recognizing humans based upon uniquely recognizing humans based upon one or more intrinsic physical or one or more intrinsic physical or behavioral traitsbehavioral traits PhysiologicalPhysiological are related to the shape of the are related to the shape of the
body, e.g., fingerprint, face recognition, body, e.g., fingerprint, face recognition, DNA, hand and palm geometry, iris DNA, hand and palm geometry, iris recognition, which has largely replaced recognition, which has largely replaced retina, and odor/scentretina, and odor/scent
BehavioralBehavioral are related to the behavior of a are related to the behavior of a person, e.g., typing rhythm, gait, and voiceperson, e.g., typing rhythm, gait, and voice
Computer Vision vs. Computer Vision vs. BiometricsBiometrics
Biometrics is a branch of Computer Biometrics is a branch of Computer VisionVision
The development of Biometrics The development of Biometrics depends on Computer Vision depends on Computer Vision techniquestechniques
Computer Vision vs. Computer Vision vs. Machine LearningMachine Learning
Machine learningMachine learning is a scientific is a scientific discipline that is concerned with the discipline that is concerned with the design and development of algorithms design and development of algorithms that allow computers to change behavior that allow computers to change behavior based on data, such as from sensor data based on data, such as from sensor data or databases (from Wikipedia)or databases (from Wikipedia)
A major focus of machine learning A major focus of machine learning research is to automatically learn to research is to automatically learn to recognize complex patterns and make recognize complex patterns and make intelligent decisions based on data. intelligent decisions based on data.
Computer Vision vs. Computer Vision vs. Machine LearningMachine Learning
Machine Learning is very useful for Machine Learning is very useful for Computer Vision (e.g., learning for vision)Computer Vision (e.g., learning for vision)
Computer Vision is more than just learningComputer Vision is more than just learning ModelingModeling Example based learningExample based learning
In Machine Learning, it usually does not In Machine Learning, it usually does not care about how to obtain the data or sensorscare about how to obtain the data or sensors
In Computer Vision, we care how to obtain In Computer Vision, we care how to obtain the visual data (sensor design, active vision), the visual data (sensor design, active vision), how to represent the visual data, and othershow to represent the visual data, and others
VisionVision
Vision is the process of discovering Vision is the process of discovering what is present in the world and what is present in the world and where it is by looking.where it is by looking.
Computer VisionComputer Vision
Computer Vision is the study of Computer Vision is the study of analysis of pictures and videos in analysis of pictures and videos in order to achieve results similar to order to achieve results similar to those as by people.those as by people.
Why Computer VisionWhy Computer Vision
An image is worth 1000 wordsAn image is worth 1000 words Many biological systems rely on Many biological systems rely on
visionvision The world is 3D and dynamicThe world is 3D and dynamic Cameras and computers are cheapCameras and computers are cheap ……
Computer Vision Computer Vision ExamplesExamples
Finding People in imagesFinding People in images
Problem 1:Problem 1: Given an image I Given an image I
Question:Question: Does I contain an image of a Does I contain an image of a person?person?
““Yes” InstancesYes” Instances
““No” InstancesNo” Instances
Some Computer Vision Some Computer Vision TopicsTopics
Imaging GeometryImaging Geometry
Camera ModelingCamera Modeling
Pinhole CamerasPinhole Cameras LensesLenses Camera Camera
Parameters and Parameters and CalibrationCalibration
Image Filtering and Image Filtering and EnhancingEnhancing
Linear Filters Linear Filters and Convolutionand Convolution
Image Image SmoothingSmoothing
Edge DetectionEdge Detection PyramidsPyramids
Image Filtering and Image Filtering and Enhancing (cont.)Enhancing (cont.)
Region SegmentationRegion Segmentation
ColorColor
TextureTexture
Image RestorationImage Restoration
Original Synthetic
Perceptual OrganizationPerceptual Organization
Perceptual OrganizationPerceptual Organization
Shape AnalysisShape Analysis
StereoStereo
Motion and Optical FlowMotion and Optical Flow
High Level VisionHigh Level Vision
Image MosaicImage Mosaic
One Very Successful One Very Successful ExampleExample
Face detection in a digital cameraFace detection in a digital camera The camera detects faces in a scene The camera detects faces in a scene
and then automatically focuses (AF) and and then automatically focuses (AF) and optimizes exposure (AE) and, if needed, optimizes exposure (AE) and, if needed, flash output. flash output.
OutlineOutline
IntroductionsIntroductions Administrative MattersAdministrative Matters Course OutlineCourse Outline Applications of Computer VisionApplications of Computer Vision Computer Vision FocusComputer Vision Focus Computer Vision PublicationsComputer Vision Publications
ApplicationsApplications autonomous cars, planes, missiles, robots, autonomous cars, planes, missiles, robots,
...... space explorationspace exploration aid to the blind, ASL recognitionaid to the blind, ASL recognition manufacturing, quality controlmanufacturing, quality control surveillance, security, biometricssurveillance, security, biometrics image retrievalimage retrieval medical imaging and analysismedical imaging and analysis ......
OutlineOutline
IntroductionsIntroductions Administrative MattersAdministrative Matters Course OutlineCourse Outline Applications of Computer VisionApplications of Computer Vision Computer Vision FocusComputer Vision Focus Computer Vision PublicationsComputer Vision Publications
Computer Vision focuses Computer Vision focuses on:on:
What information should be What information should be extracted?extracted?
How can it be extracted?How can it be extracted? How should it be represented?How should it be represented? How can it be used to achieve the How can it be used to achieve the
goal?goal?
Related disciplinesRelated disciplines
Image processingImage processing Pattern recognitionPattern recognition PhotogrammetryPhotogrammetry Computer graphicsComputer graphics Artificial intelligenceArtificial intelligence Machine learningMachine learning Projective geometryProjective geometry Control theoryControl theory
Active Research TopicsActive Research Topics
Object recognitionObject recognition Human behavior analysisHuman behavior analysis Internet and computer visionInternet and computer vision Biometrics and soft biometricsBiometrics and soft biometrics Large scale 3D reconstruction (city level)Large scale 3D reconstruction (city level) Medical image processing Medical image processing Vision for roboticsVision for robotics ……
OutlineOutline
IntroductionsIntroductions Administrative MattersAdministrative Matters Course OutlineCourse Outline Applications of Computer VisionApplications of Computer Vision Computer Vision FocusComputer Vision Focus Computer Vision PublicationsComputer Vision Publications
Computer Vision Computer Vision PublicationsPublications
JournalsJournals IEEE Trans. on Pattern Analysis and Machine IEEE Trans. on Pattern Analysis and Machine
Intelligence (TPAMI)Intelligence (TPAMI) #1 IEEE, Thompson-ISI impact factor: 5.96#1 IEEE, Thompson-ISI impact factor: 5.96 #1 in both electrical engineering and artificial #1 in both electrical engineering and artificial
intelligenceintelligence #3 in all of computer science#3 in all of computer science
Internal Journal of Computer Vision (IJCV)Internal Journal of Computer Vision (IJCV) ISI impact factor: 5.358, Rank 2 of 94 in “CS, artificial ISI impact factor: 5.358, Rank 2 of 94 in “CS, artificial
intelligenceintelligence IEEE Trans. on Image ProcessingIEEE Trans. on Image Processing ……
Importance of CVImportance of CV
From these major journal rankings, From these major journal rankings, we can see the importance of we can see the importance of Computer Vision research in the Computer Vision research in the whole areas of whole areas of Computer ScienceComputer Science Electrical EngineeringElectrical Engineering
Computer Vision Computer Vision PublicationsPublications
ConferencesConferences International Conference on Computer International Conference on Computer
Vision (ICCV)Vision (ICCV) Conf. of Computer Vision and Pattern Conf. of Computer Vision and Pattern
Recognition (CVPR)Recognition (CVPR) Europe Conference on Computer Vision Europe Conference on Computer Vision
(ECCV)(ECCV) ……
Discussions and Discussions and QuestionsQuestions