計算機視覺教程

計算機視覺課程

COMPUTERVISION

Referencebooks:1.MachineVision,byRamerhJian,etal(MajorReference)2.ComputerVision,byDamaH.Ballard,etal.3.Vision,ByDavidMarr.4.IntroductoryTechniquesfor3-DComputerVision,byEmanueleTrucco.5.ComputerVisionandImageProcessing,byScatt

E.Umbaugh.6.AlgorithmforImageProcessingandComputerVision,by

.7.機器視覺，賈云得編著8.計算機視覺的理論和實踐，李介谷著9.計算機視覺——計算機理論與算法根底，馬頌德Ruleofgrading:

Homework: 10%

Examination: 60%

Project: 30%

ImportantDeadlines:

Handinprojectproposal:May11,2021

Handinprojectreport:July6,2021

AllwritteninChinese(中文書寫)

FinalExamination:July6,2021Course1 IntroductionCourse1 Introduction1.WhatisComputerVisionComputervisionaimstorecoverusefulinformationabouta(3D)scenefromits2Dprojections(images).——3Ddepth/3Dstructure——3Dmotionanalysis——3Dsurfacesandorientation——statusof3Dobjects,soon…——meaningofactionsof3Dobjects(cognition)Everytaskistoughsinceitistobeunderstoodfromflatimages!

Computervisionisalsocalledas:

Machinevision

Robotvision

Imageunderstanding

Sceneanalysis(inearlyyears)課程內(nèi)容：

1)

緒論(Introduction)

2)

圖象濾波(ImageFiltering)

3)

二進(jìn)制圖象處理(BinaryImageProcessing)

4)

區(qū)域(Region)

5)

邊緣檢測(EdgeDetection)

6)

立體視覺(Stereo)

7)

運動的理解與估值(Motion)

8)

輪廓(Contours)

9)

紋理(Texture)

10)

圖象光度學(xué)(Shading)

11)

光流場(OpticFlows)

12)

系統(tǒng)校準(zhǔn)(Calibration)

13)

曲線與曲面(CurvesandSurfaces)

14)

動態(tài)視覺(DynamicVision)

15)三維識別(ObjectRecognition)2.Relatedfields

——imageprocessing

——computervision

——patternrecognition

——artificialIntelligence

——neuralnetwork

——psychophysics

3.PerspectiveProjection

——pin-holecameramodel

projectioncenter(i.e.,camera)isattheoriginofscenecoordinatesystemwithopticaxisofcamerapointingtoz-axisdirectionofcoordinatesoriginofimagecoordinatesystemislocatedat(0,0,f)ofscenecoordinatesystem(fisfocuslengthofthecamera),theX-andY-axesofimagecoordinatesareparallelwithx-andy-axesofscenecoordinates,respectivelycameradoesnothaveanylensdistortions.xoyoXYP(X,Y)P(x,y,z)LlnP0focuslength=frRz(1)Imagepoint

ImageLine

3DlinelisprojectedontoimageplanethroughprojectingplanetogetimagelineL:

Letweuseavector

torepresenttheimagelineLofl.SinceWehave,orItistheequationofprojectingplane,thenormalis.Thus,wewillusethenormaloftheprojectingplaneof3Dlinetodenoteimageplaneby4.DigitalImageSampling:digitalizeinimageplane(directions)

say,,,andsoon.quantization:digitalizeinima