Downloaded 16 times
Lecture 05: Vision
- 1. Introduction to RoboticsPerceptionIICSCI 4830/7000February 15, 2010NikolausCorrell
- 2. Review: SensingImportant: sensorsreport data in their own coordinate frameExamples from the exerciseAccelerometer of NaoLaser scannerTreat like forward kinematics
- 3. TodayPerception using visionPracticalangle:Why is vision hardBasic image processingHow to combine image processing primitives into object recognitionOpenCV / SwisTrack
- 4. Why is VisionHard?The difference between seeing and perception.Gary Bradski, 20094What to do? Maybe we should try to find edges ….Gary Bradski, 2005
- 5.
- 6.
- 7. Reflectance discontinuity (i.e.,change in surface material properties)
- 8. Illumination discontinuity (e.g.,shadow)Slide credit: Christopher Rasmussen5But, What’s an Edge?
- 9. To Deal Withthe Confusion, Your Brain has Rules...That can be wrong
- 10. We even seeinvisible edges…
- 11.
- 12. We need todeal with 3D Geometry9Perception is ambiguous … depending on your point of view!Graphic by Gary Bradski
- 13. And Lighting in3DWhich square is darker?
- 14. Lighting is Ill-posed…Perception of surfaces depends on lighting assumptions11Gary Bradski (c) 200811
- 15. Contrast12Which one ismale and which one is female?Illusion by: Richard Russell,Harvard UniversityRussell, R. (2009) A sex difference in facial pigmentation and its exaggeration by cosmetics. Perception, (38)1211-1219
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22. Thin LensObjects needto have the right distance to be in focus -> Depth-from-Focus method
- 23.
- 24. Canny Edge Detector21GaryBradski (c) 200821
- 25. Morphological Operations ExamplesMorphology- applying Min-Max. Filters and its combinationsDilatation IBOpening IoB= (IB)BErosion IBImage IClosing I•B= (IB)BTopHat(I)= I - (IB)BlackHat(I)= (IB) - IGrad(I)= (IB)-(IB)
- 26. Stereo CalibrationScreen shots and charts by Gary Bradski, 2005Gary Bradski (c) 20082323
- 27. 3D Stereo VisionFindEpipolar lines:Align images:Triangulate points:Depth:
- 28. Example: Tomato-Picking RobotChallengesFoliageReflectionsVaryingsize and shapeVarying colorPartly covered fruitshttp://swistrack.sourceforge.netN. Correll, N. Arechiga, A. Bolger, M. Bollini, B. Charrow, A. Clayton, F. Dominguez, K. Donahue, S. Dyar, L. Johnson, H. Liu, A. Patrikalakis, T. Robertson, J. Smith, D. Soltero, M. Tanner, L. White, D. Rus. Building a Distributed Robot Garden. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 1509-1516, St. Louis, MO.
- 29. Filter-based object recognitionFilterimageSobelHough transformColorSpectral highlightsSize and shapeWeighted sum of filters highlights object locationColorSobelHoughSpectralHighlights
- 30.
- 31. HomeworkRead sections 4.2-5(pages 145-180)Questionnaire on CU Learn