블로그 이미지
Leeway is... the freedom that someone has to take the action they want to or to change their plans.
maetel

Notice

Recent Post

Recent Comment

Recent Trackback

Archive

calendar

1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30
  • total
  • today
  • yesterday

Category

'글타래'에 해당되는 글 750건

  1. 2010.11.11 Velho, Frery & Gomes [Image Processing for Computer Graphics and Vision] (2nd ed)
  2. 2010.11.11 Duda, Hart & Stork [Pattern Classification] (2nd ed)
  3. 2010.11.02 OpenCV 2.1.0 Installation on Mac OS X Snow Leopard
  4. 2010.11.01 Camera 4 Line Art By darjjeelling
  5. 2010.10.31 Pattie Maes and Pranav Mistry demo SixthSense
  6. 2010.10.25 CAPTIG: On-line Conference Call
  7. 2010.10.18 Ian Piper <Learn Xcode Tools for Mac OS X and iPhone Development>
  8. 2010.10.16 추천서: 데이브 마크 & 제프 라마시 <아이폰 3 프로그래밍> (위키북스, 2009)
  9. 2010.10.16 ISMAR 2010: B O R D E R L E S S
  10. 2010.10.12 OpenCV: Decision Trees
  11. 2010.10.12 OpenCV: Random Trees
  12. 2010.10.10 1강: iPhone Framework
  13. 2010.10.07 [오재혁] 아이폰 어플리케이션 제작 워크숍: Preparation and Practice
  14. 2010.10.07 David M. L. Williams 서강대 영상대학원 특강
  15. 2010.10.02 Object Detection in Crowded Workspaces
  16. 2010.10.01 variational principle
  17. 2010.09.27 OpenCV: cvFindChessboardCorners() 함수
  18. 2010.09.26 OpenCV: cvThreshold() 함수 연습
  19. 2010.09.26 Matthias Felleisen et al. <How to Design Programs>
  20. 2010.09.26 Dazhi Chen & Guangjun Zhan, "A New Sub-Pixel Detector for X-Corners in Camera Calibration Targets"
  21. 2010.09.26 Luca Lucchese & Sanjit K. Mitra "Using saddle points for subpixel feature detection in camera calibration targets"
  22. 2010.09.25 D. C. Brown,
  23. 2010.09.25 Otsu's method
  24. 2010.09.24 Janne Heikkila & Olli Silven "A Four-step Camera Calibration Procedure with Implicit Image Correction"
  25. 2010.09.24 Learning OpenCV: Chapter 11 Camera Models and Calibration
2010. 11. 11. 16:22 Computer Vision
Image Processing for Computer Graphics and Vision, 2nd ed. 
Velho, Luiz, Frery, Alejandro C., Gomes, Jonas
Springer, 2009


비록 학부 4학년은 아니었더라도 석사 1학기에 봤었다면 좋았을 것을...


'Computer Vision' 카테고리의 다른 글

Winsock  (0) 2010.11.26
camera Firefly (FFMV-03M2M)  (0) 2010.11.25
Duda, Hart & Stork [Pattern Classification] (2nd ed)  (0) 2010.11.11
OpenCV 2.1.0 Installation on Mac OS X Snow Leopard  (0) 2010.11.02
OpenCV: Decision Trees  (0) 2010.10.12
posted by maetel
2010. 11. 11. 01:26 Computer Vision
Richard O. Duda, Peter E. Hart, David G. Stork, Pattern Classification (2nd ed-4th print), Wiley-Interscience

RICHARD O. DUDA
PhD, Professor in the Electrical Engineering Department at San Jose State University, San Jose, California.

PETER E. HART
PhD, Chief Executive Officer and President of Ricoh Innovations, Inc. in Menlo Park, California.

PhD, Chief Scientist, also at Ricoh Innovations, Inc.

lectures: 



cf.
Computer Manual in MATLAB to accompany Pattern Classification, 2nd EditionDavid G. Stork ( Ricoh Silicon Valley ), Elad Yom-Tov, John Wiley & Sons, 2004
loyola: 1관 4층 006.4 S885c 2004

DAVID G. STORK
PhD, Chief Scientist at Ricoh Innovations, Inc., and Consulting Professor of Electrical Engineering at Stanford University. A graduate of MIT and the University of Maryland, he is the founder and leader of the Open Mind Initiative and the coauthor, with Richard Duda and Peter Hart, of Pattern Classification, Second Edition, as well as four other books.

PhD, research scientist at IBM Research Lab in Haifa, working on the applications of machine learning to search technologies, bioinformatics, and hardware verification (among others). He is a graduate of Tel-Aviv University and the Technion.



Preface


"(Our purpose is) to give a systematic account of the major topics  in pattern recognition, based on fundamental principles"

pattern recognition

speech recognition
optical character recognition
signal classification

pattern classification
scene analysis
machine learning

handwriting & gesture recognition
lipreading
geological analysis
document searching
recognition of bubble chamber tracks of subatomic particles
human-machine interface - eg. pen-based computing
human and animal nervous systems

neurobiology
psychology

"We address a specific class of problems - pattern recognition problems - and consider the wealth of different techniques that can be applied to it."

"We discuss the relative strengths and weaknesses of various classification techniques"


statistical methods vs. syntactic methods





posted by maetel
2010. 11. 2. 21:53

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

2010. 11. 1. 01:34 Cases/apps
사진을 찍으면 edges만을 살려 크로키 느낌을 주는 앱
http://itunes.apple.com/app/camera-4-line-art/id398988727?mt=8





풍경화 느낌을 주는 Camera 4 Landspace도 있다.
http://itunes.apple.com/app/camera-4-landscape/id391360591?mt=8

posted by maetel
2010. 10. 31. 23:41 Cases

'Cases' 카테고리의 다른 글

David Reed  (0) 2012.06.29
Object Detection in Crowded Workspaces  (0) 2010.10.02
Bayard and Eco: How to Talk About Books You Haven't Read  (0) 2010.09.18
Richard Dreyfuss: Improving Civic Education  (0) 2010.08.04
Oyster card  (0) 2008.08.14
posted by maetel
2010. 10. 25. 12:23

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

2010. 10. 18. 23:36 Method/IDE

Learn Xcode Tools for Mac OS X and iPhone Development
by Ian Piper
google books: http://books.google.com/books?id=CwjLlUDFoLkC


번역서: 맥과 아이폰 개발자를 위한 XCODE   
이안 파이퍼  저 | 황반석 역 | 제이펍 | 2010.03.29
네이버 책: http://book.naver.com/bookdb/book_detail.nhn?bid=6269096





* Memory Leak
http://books.google.com/books?id=CwjLlUDFoLkC&lpg=PP1&hl=ko&pg=PA229#v=onepage&q&f=false

'Method > IDE' 카테고리의 다른 글

MathWorks Korea 세미나 2009년 11월  (0) 2009.11.19
Peter Elsea <Visualization of Audio>  (0) 2006.12.11
posted by maetel
번역서: 시작하세요! 아이폰 3 프로그래밍: IPHONE SDK를 이용한 아이폰 개발

원서: More IPhone 3 Development: Tackling IPhone SDK 3  by Dave Mark, Jeff LaMarch
번역서: MORE 아이폰 3 프로그래밍   (IPHONE SDK 3 집중분석)


http://iphonedevbook.com/


posted by maetel
2010. 10. 16. 12:11 Footmarks
http://www.ismar10.org



2010-10-14 목





2010-10-15 금

Jeffry Shaw
Artevertiser - Improved Reality





posted by maetel
2010. 10. 12. 13:14 Computer Vision
posted by maetel
2010. 10. 12. 13:09 Computer Vision

'Computer Vision' 카테고리의 다른 글

OpenCV 2.1.0 Installation on Mac OS X Snow Leopard  (0) 2010.11.02
OpenCV: Decision Trees  (0) 2010.10.12
variational principle  (0) 2010.10.01
OpenCV: cvFindChessboardCorners() 함수  (0) 2010.09.27
OpenCV: cvThreshold() 함수 연습  (0) 2010.09.26
posted by maetel
2010. 10. 10. 00:52

보호되어 있는 글입니다.
내용을 보시려면 비밀번호를 입력하세요.

아이폰 어플리케이션 제작 워크숍: Preparation and Practice
강사: 오재혁
시간: 10년 10월 8일부터 총 10회, 금요일 오후 7시 30분 ~9시 30분
수강료: 300,000원



강좌 개요
iPhone의 매력은 사용 경험에 기반한다. 일상의 한 부분에 퍼즐 조각처럼 들어맞는, 어플리케이션을 사용하며 얻는 만족감은, 주변에서 쉽게 확인할 수 있다. 하지만, iPhone 플랫폼의 매력은 사용에 한정되지 않는다. iPhone 어플리케이션 제작은 더욱 농밀하고 치명적인 경험 기회를 제공한다. 직접 제작한, 자신만의 표현을 간직한 어플리케이션을, 전 세계에 배포할 수 있고, 그것을 매개로 소통할 수 있다는 가능성은, 제작자가 되기 위한 노력을 상쇄하고 남을 만큼의 가치를 제공한다.

본 강좌는 iPhone 어플리케이션 제작의 경험을 같이 하고자 하는 사용자를 대상으로 한다. 어플리케이션 제작을 위해 알아야 할 지식들을 살펴보고, 전반적인 기능을 연습해본다.

* 필요한 준비 사항
: 수강을 위해서는 Snow Leopard (Mac OS X 10.6.x) 이후 버전이 설치된 Mac 과 iPod Touch 혹은 iPhone이 필요합니다.



강사 소개
오재혁 / 프로그래머
서울대학교에서 컴퓨터공학을 전공하였으며 다수의 인터랙티브 설치 작업을 제작 및 지원해왔다. 현재 프리랜서로 설치 제작 및 team Gurunun 에서 아이폰 어플리케이션 제작에 주력하고 있다.



강좌 계획

1강: iPhone Framework
- iPhone SDK 내용 전반을 훑어본다.
- Xcode (어플리케이션 제작 환경) 사용법을 익힌다.
- SDK 에서 제공하는 template project 에서 간단한 어플리케이션을 만들고, 실행해본다.

2강:Objective-C, Cocoa
- Objective-C, Cocoa 의 역사, 특징을 이해한다.
- iPhone Application Project의 구성을 살펴본다.
- iPhone SDK 에서 원하는 부분을 찾아 사용하는 법을 배운다.

3강: Primitive Interface
- 기본적인 UI 컨트롤 (Button, TextField, Slider, Switch, Progress View, … ) 사용법을 익힌다.

4강: Advanced Interface
- Table View, Navigation Controller 를 사용하여 hierarchical list 를 구성해본다.

5강: Animation
- Core Animation 을 활용하여 UI 컨트롤을 움직여 본다.
- 보다 견고한 Animation 진행을 위한 State Machine 을 구성해본다.

6강: Quartz 2D, OpenGL ES
- Quartz 2D 를 사용한 그리기를 연습한다.
- OpenGL ES template project 를 분석하고, 3D Drawing 의 구성 요소를 이해한다.

7강: Touch, Accelerometer
- Accelerometer 값을 받아들여, 해석, 사용하는 법을 연습한다.
- Touch 입력을 처리하고, Gesture Recognizer 의 구조와 사용법을 이해한다.

8강: Audio
- iPod library 의 음악을 연주하는 법을 알아본다.
- 여러 음원을 동시에 출력하는 법을 알아본다.
- OpenAL 의 기능을 이해하고, 사용해본다.

9강: Advanced Audio
- Audio Queue 를 활용하여, linear PCM 을 출력해본다.
- Microphone 입력을 받아들이는 방법을 알아본다.

10강: Connection
- 웹의 데이터를 읽어들이는 방법을 연습해본다.
- GameKit 을 활용하여 peer-to-peer 통신을 시도해본다.



posted by maetel
2010. 10. 7. 23:00 Footmarks
2010-10-07 @서강대 영상대학원 가브리엘관 703호

David M. L. Williams's bio:
born in London
B.S.  in Information System
Ph. D. Cognitive Science
UX
HCI

Mojo Interactive Spaces Mojoispaces.com

posted by maetel
2010. 10. 2. 07:30 Cases

'Cases' 카테고리의 다른 글

David Reed  (0) 2012.06.29
Pattie Maes and Pranav Mistry demo SixthSense  (0) 2010.10.31
Bayard and Eco: How to Talk About Books You Haven't Read  (0) 2010.09.18
Richard Dreyfuss: Improving Civic Education  (0) 2010.08.04
Oyster card  (0) 2008.08.14
posted by maetel
2010. 10. 1. 18:42 Computer Vision
posted by maetel
2010. 9. 27. 01:47 Computer Vision
OpenCV: cvFindChessboardCorners or cv::findChessboardCorners

source code: https://code.ros.org/trac/opencv/browser/tags/2.1/opencv/src/cv/cvcalibinit.cpp
file:///opencv/src/cv/cvcalibinit.cpp


ref.
Learning OpenCV: Chapter 11 Camera Models and Calibration

V.Vezhnevets, A.Velizhev (Graphics and Media Lab, CMC department, Moscow State University) "GML C++ Camera Calibration Toolbox", 2005
http://graphics.cs.msu.ru/en/science/research/calibration/cpp


http://www.vision.caltech.edu/bouguetj/calib_doc/htmls/example.html


IplImage* cvGetImage(const CvArr* arr, IplImage* imageHeader)


cvCheckChessboard
source file: //opencv/src/cv/cvcheckchessboard.cpp

Dilate or cv::dilate

Flood-fill algorithm
http://en.wikipedia.org/wiki/Flood_fill


quad_count = icvGenerateQuads( &quads, &corners, storage, thresh_img, flags );


CvMemStorage* cvCreateChildMemStorage(CvMemStorage* parent)

StartFindContours

cvContourArea






routine:
"cvcalibinit.cpp": cvFindChessboardCorners()'s routine

1. gray-scale input image (option: CV_CALIB_CB_NORMALIZE_IMAGE | CV_CALIB_CB_NORMALIZE_IMAGE ) -> save "imgGray"


2. binarize the gray-scaled image

adative thresholding (option: CV_CALIB_CB_ADAPTIVE_THRESH) or thresholding by "image mean-10" or "10"




3. dilate the binarized image -> save "imgThresh"





4. find rectangles to draw white lines around the image edge -> save "imgRect"

 


1. check if a chessboard is in the input image

 

1) erode and dilate ( cvErode() & cvDilate() )

2) find a threshold value to make contours ( "flag" )
3) select contours to make quadrangles
4) check if there are many hypotheses with similar sizes ( floodfill

-style algorithm )

 

2. (if pattern was no found using binarization) multi-level quads

extraction


3. draw white lines around the image edges ( "thresh_image" )


4. compute corners in clockwise order


5. find the quadrangle's neighbors


6. find connected quadrangles to order them and the corners


7. remove extra quadrangles to make a nice square pattern


8. check if each row and column of the chessboard is monotonous


9. refine corner locations








1. check if a chessboard is in the input image

1) erode and dilate ( cvErode() & cvDilate() )
2) find a threshold value to make contours ( "flag" )
3) select contours to make quadrangles
4) check if there are many hypotheses with similar sizes ( floodfill-style algorithm )
2. (if pattern was no found using binarization) multi-level quads extraction
3. draw white lines around the image edges ( "thresh_image" )
4. compute corners in clockwise order
5. find the quadrangle's neighbors
6. find connected quadrangles to order them and the corners
7. remove extra quadrangles to make a nice square pattern
8. check if each row and column of the chessboard is monotonous
9. refine corner locations



 

 
 

posted by maetel
2010. 9. 26. 19:30 Computer Vision
OpenCV 함수 cvTreshold or cv::threshold

source code link: https://code.ros.org/trac/opencv/browser/tags/2.1/opencv/src/cv/cvthresh.cpp
file: /opencv/src/cv/cvthresh.cpp


Learning OpenCV: Chapter 5 Image Processing: Threshold (135p)


cf. Otsu method




원본 영상

입력 영상



threshold=100, max_val=100, CV_THRESH_BINARY

threshold=100, max_val=200, CV_THRESH_BINARY

threshold=200, max_val=200, CV_THRESH_BINARY



threshold=50, CV_THRESH_TRUNC

threshold=50, CV_THRESH_TRUNC

threshold=150, CV_THRESH_TRUNC



threshold=50, CV_THRESH_TOZERO

threshold=100, CV_THRESH_TOZERO

threshold=200, CV_THRESH_TOZERO



threshold=100, max_val=200, CV_THRESH_BINARY

threshold=100, max_val=200, CV_THRESH_BINARY & CV_THRESH_OTSU

threshold=100, max_val=200, CV_THRESH_OTSU


코드에서 CV_THRESH_OTSU는 CV_THRESH_BINARY | CV_THRESH_OTSU 와 같은 효과.
CV_THRESH_OTSU는 함수의 인자 "threshold"의 초기값과 무관하게 입력 영상에 대해 내부적으로 threshold 값을 구하고 이에 따라 선택된 픽셀들에 max_value 값을 준다.


posted by maetel
2010. 9. 26. 19:18 Computation
How to Design Programs        
: An Introduction to Computing and Programming           

http://www.htdp.org/

Matthias Felleisen        
Robert Bruce Findler        
Matthew Flatt        
Shriram Krishnamurthi           

The MIT Press        
Cambridge, Massachusetts        
London, England

pdf download

'Computation' 카테고리의 다른 글

안드로이드 내부구조 #1  (0) 2011.10.24
매킨토시 스크린 샷 파일 포맷 변경법  (0) 2011.02.10
Octave on Mac OS X  (1) 2009.08.12
Cloud Computing 클라우드 컴퓨팅  (0) 2009.07.03
OpenFrameworks  (2) 2008.06.18
posted by maetel
2010. 9. 26. 15:57 Computer Vision
Dazhi Chen & Guangjun Zhan, "A New Sub-Pixel Detector for X-Corners in Camera Calibration Targets," WSCG Short Papers (2005): 97-100



The School of Instrumentation Scienc & Optoelectronics Engineering (SISOE), Beijing University of Aeronautics and Astronautics
posted by maetel
2010. 9. 26. 15:47 Computer Vision
L. Lucchese and S.K. Mitra. Using saddle points for subpixel feature detection in camera calibration targets. In Proceedings of the 2002, Asia Pacific Conference on Circuits and Systems, volume 2, pages 191-195, 2002.



posted by maetel
2010. 9. 25. 15:42 Computer Vision

D. C. Brown,




Duane C. Brown Award
http://www.lsgi.polyu.edu.hk/staff/Bo.Wu/duane_c._brown_awards_description.htm

PCV @ OSU (Photogrammetric Computer Vision Laboratory, Ohio State University)

 http://academic.research.microsoft.com/Author/5813842.aspx


D. C. Brown, "Decentering distortion of lenses", Photogrammetric Engineering 32(3) (1966). 7: 444–462.








posted by maetel
2010. 9. 25. 01:44 Computer Vision
http://en.wikipedia.org/wiki/Otsu's_method informed by prof.


Nobuyuki Otsu (1979). "A threshold selection method from gray-level histograms". IEEE Trans. Sys., Man., Cyber. 9: 62–66. doi:10.1109/TSMC.1979.4310076



OpenCV 함수 cvThreshold()


ref.
Bryan S. Morse's Brigham Young University (1998-2000) Lecture 4: Thresholding

Milan Sonka, Vaclav Hlavac, Roger Boyle, <Image Processing, Analysis, and Machine Vision> (3rd ed.), Thomson (2008)
: Chapter 6 Segmentation I: 6.1.2 Optimal thresholding (180p)






posted by maetel
2010. 9. 24. 23:16 Computer Vision
Janne Heikkila, Olli Silven, "A Four-step Camera Calibration Procedure with Implicit Image Correction," cvpr, pp.1106, 1997 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'97), 1997


Matlab toolbox: http://www.ee.oulu.fi/~jth/calibr/

JanneHeikkilä - Machine Vision Group, Computer Science and Engineering Laboratory, University of Oulu
Olli Silven - Information Processing Laboratory, University of Oulu

cf.
http://www.vision.caltech.edu/bouguetj/calib_doc/htmls/parameters.html


1. Introduction

4-step
1) DLT (direct linear transformation) for initial parameter values [Tsai 1987, Abdel-Aziz 1971, Melen 1994]
2) Nonlinear parameter estimation for final parameter values [Slama 1980]
3) Correcting errors sourced from feature extraction
4) image correction: new implicit model to interpolate the correct image points


2. Explicit camera calibration

The pinhole camera model is based on the principle of collinearity, where each point in the object space is projected by a straight line through the projection center into the image plane.

http://en.wikipedia.org/wiki/Euler_angle











posted by maetel
2010. 9. 24. 16:24 Computer Vision
Learning OpenCV ebook
: Chapter 11 Camera Models and Calibration


ref.
opencv v2.1 documentation » cv. Image Processing and Computer Vision » Camera Calibration and 3D Reconstruction (c reference / cpp reference)
Noah Kuntz's OpenCV Tutorial 10 - Chapter 11


370p
detection of light from the world:
light source -> reflected light from a object -> our eye or camera (lens -> retina or imager)


cf. http://en.wikipedia.org/wiki/Electronic_imager


geometry of the ray's travel
 
pinhole camera model

ref. O'Connor 2002
al-Hytham (1021)
Descartes
Kepler
Galileo
Newton
Hooke
Euler
Fermat
Snell

ref.
Trucco 1998
Jaehne 1995;1997
Hartley and Zisserman 2006
Forsyth and Ponce 2003
Shapiro and Stockman 2002
Xu and Zhang 1996


projective geometry

lens distortion

camera calibration
1) to correct mathematically for the main deviations from the simple pinhole model with lenses
2) to relate camera measurements with measurements in the real, 3-dimensional world


3-D scene reconstruction


: Camera Model (371p)

camera calibration => model of the camera's geometry & distortion model of the lens : intrinsic parameters


homography transform

OpenCV function cvCalibrateCamera2() or cv::calibrateCamera

pinhole camera:
single ray -> image plane (projective plane)

(...) the size of the image relative to the distant object is given by a single parameter of the camera: its focal length. For our idealized pinhole camera, the distance from the pinhole aperture to the screen is precisely the focal length.

The point in the pinhole is reinterpreted as the center of projection.

The point at the intersection of the image plane and the optical axis is refereed to as the principal point.

(...) the individual pixels on a typical low-cost imager are rectangular rather than square. The focal length(fx) is actually the product of the physical focal length(F) of the lens and the size (sx) of the individual imager elements. (*sx converts physical units to pixel units.)


: Basic Projective Geometry (373p)

projective transform

homogeneous coordinates

The homogeneous coordinates associated with a point in a projective space of dimension n are typically expressed as an (n+1)-dimensional vector with the additional restriction that any two points whose values are proportional are equivalent.

camera intrinsics matrix (of parameters defining our camera, i.e., fx,fy,cx,and cy)

ref. Heikkila and Silven 1997


OpenCV function cvConvertPointsHomogeneous() or cv::convertPointsHomogeneous

cf. cvReshape or cv::reshape

For a camera to form images at a faster rate, we must gather a lot of light over a wider area and bend (i.e., focus) that light to converge at the point of projection. To accomplish this, we uses a lens. A lens can focus a large amount of light on a point to give us fast imaging, but it comes at the cost of introducing distortions.


: Lens Distortions (375p)

Radia distortions arise as a result of the shape of lens, whereas tangential distortions arise from the assembly process of the camera as a whole.


radial distortion:
External points on a frontfacing rectangular grid are increasingly displaced inward as the radial distance from the optical center increases.

"barrel" or "fish-eye" effect -> barrel distortion


tangential distortion:
due to manufacturing defects resulting from the lens not being exactly parallel to the imaging plane.


plumb bob (연직 추, 측량 추) model
ref. D. C. Brown, "Decentering distortion of lenses", Photogrammetric Engineering 32(3) (1966). 7: 444–462.
http://www.vision.caltech.edu/bouguetj/calib_doc/htmls/parameters.html

distortion vector (k1, k2, p1, p2, k3); 5-by-1 matrix


: Calibration (378p)

ref. http://www.vision.caltech.edu/bouguetj/calib_doc/

cvCalibrateCamera2() or cv::calibrateCamera
The method of calibration is to target the camera on a known structure that has many individual and identifiable points. By viewing this structure from a variety of angles, it is possible to then compute the (relative) location and orientation of the camera at the time of each image as well as the intrinsic parameters of the camera.


얘네는 뭔 있다 없다 하니...

cv::initCameraMatrix2D

cvFindExtrinsicCameraParams2



: Rotation Matrix and Translation Vector (379p)

Ultimately, a rotation is equivalent to introducing a new description of a point's location in a different coordinate system.


Using a planar object, we'll see that each view fixed eight parameters. Because the six parameters on two additional parameters that we use to resolve the camera intrinsic matrix. We'll then need at least two views to solve for all the geometric parameters.


: Chessboards (381p)

OpenCV opts for using multiple view of a planar object (a chessboard) rather than one view of  a specially constructed 3D object. We use a pattern of alternating black and white squares, which ensures that there is no bias toward one side or the other in measurement. Also, the resulting gird corners lend themselves naturally to the subpixel localization function.

use a chessboard grid that is asymmetric and of even and odd dimensions - for example, (5,6). using such even-odd asymmetry yields a chessboard that has only one symmetry axis, so the board orientation can always be defined uniquely.


The chessboard interior corners are simply a special case of the more general Harris corners; the chessboard corners just happen to be particularly easy to find and track.







cf. Chapter 10: Tracking and Motion: Subpixel Corners
319p: If you are processing images for the purpose of extracting geometric measurements, as opposed to extracting features for recognition, then you will normally need more resolution than the simple pixel values supplied by cvGoodFeaturesToTrack().

fitting a curve (a parabola)
ref. newer techniques
Lucchese02
Chen05


cvFindCornerSubpix() or cv::cornerSubPix (cf. cv::getRectSubPix )



ref.
Zhang99; Zhang00
Sturm99


: Homography (384p)




posted by maetel