10 10 conclusions
play

10/10: Conclusions Jack Tumblin EECS, Northwestern University The - PowerPoint PPT Presentation

A Gentle Introduction A Gentle Introduction to Bilateral Filtering to Bilateral Filtering and its Applications and its Applications 10/10: Conclusions Jack Tumblin EECS, Northwestern University The Bilateral Filter The Bilateral Filter


  1. A Gentle Introduction A Gentle Introduction to Bilateral Filtering to Bilateral Filtering and its Applications and its Applications 10/10: Conclusions Jack Tumblin – EECS, Northwestern University

  2. The Bilateral Filter The Bilateral Filter • As Simple as Weights from TWO Gaussian Functions; • Has deep connections to PDE formalisms, shock-forming, heat flow, and diffusion • Speed? several very fast versions

  3. Bilateral Filter Bilateral Filter GOALS are Ambitious & Subjective, but GOALS are Ambitious & Subjective, but • Noise Removal • “Edge-Preserving Smoothing” • Locally Weighted Averaging with Outlier Rejection • Visual Separation of ‘fine details’ and ‘large features’ METHODS are Simple & Practical METHODS are Simple & Practical • Bilateral Filter: doubly-weighted local average, High weights only for ‘nearby’, ‘similar’ values • YOU set parameters of what is ‘nearby’, of what is ‘similar’

  4. VERY Broad Range of Applications: VERY Broad Range of Applications: • any task where ‘edges’ and ‘similarities’ carry useful information • data of arbitrary dimensions; M to N, time, color, image vector-spaces, etc, • even on manifolds � 3D meshes, etc.

  5. Explosive Growth: *6* New Papers Here Explosive Growth: *6* New Papers Here SIGGRAPH 2007: don’t miss these new works! . • Image Deblurring with L. Yuan, et al. Blurred/Noisy Image Pairs • Multiscale Shape and Detail Enhancement R. Fattal et al. from Multi-Light Image Collections J. Kopf et al. • Joint Bilateral Upsampling • LDR2HDR: On-the-fly Reverse Tone Mapping A. G. Rempel et al. of Legacy Video and Photographs K. Sunkavalli et al. • Factored Time-Lapse Video • Real-Time Edge-Aware Image Processing J. Chen et al. With the Bilateral Grid

  6. Is All the Work Finished? Is All the Work Finished? • Nothing left to explore? No, No, No! Bilateral filters raise deep, basic questions; Answers still vague and tentative.. For example …

  7. What Is ‘Noise’? What is ‘Texture’? What Is ‘Noise’? What is ‘Texture’? Wet Sand (Jay Sekora)

  8. What Is ‘Noise’? What is ‘Texture’? What Is ‘Noise’? What is ‘Texture’? Human Skin (Ken Perlin)

  9. What Is an ‘Edge’ in Human Vision? What Is an ‘Edge’ in Human Vision? Illusory Contours (G. Kanizsa, 1955)

  10. Does absolute intensity matter? Does absolute intensity matter? Often, Perceived Intensities ≠ Pixel Values: Example: ‘ ‘Simultaneous Contrast Simultaneous Contrast’ ’ Example: Center strip has CONSTANT CONSTANT intensity intensity… … Center strip has

  11. Does absolute intensity matter? Does absolute intensity matter? Often, Perceived Intensities ≠ Pixel Values: Example: ‘ ‘Simultaneous Contrast Simultaneous Contrast’ ’ Example: Center strip has CONSTANT CONSTANT intensity intensity… … Center strip has

  12. What Is an Illumination Edge? What Is an Illumination Edge? Sometimes it isn’t a large intensity change…

  13. What Is a ‘Geometric Edge’ in Images? What Is a ‘Geometric Edge’ in Images? 3D ‘Peanut’ shape Some silhouettes are SUGGESTED by shape cues

  14. What Is an ‘Edge’ at The Finest Scales? What Is an ‘Edge’ at The Finest Scales? Scale Problems: Can’t resolve every hair and fiber in fur… Long Black Whiskers: • Visible on Bkgnd, but • Visible against fur? Albrecht Durer (1502)

  15. For any perceivable image? For any perceivable image? Hmm. Edge? Noise? Regions? Texture? Silhouette? … http:// mfrost.typepad.com/cute_overload/2007/02/backlit_extrava.html

  16. Thank You For Attending! Thank You For Attending! • Please do fill out a review form, and tell us how we can improve the course… http://www.siggraph.org/courses_evaluation • For more in-depth info: links, and all docs: http://people.csail.mit.edu/sparis/siggraph07_course/

  17. Course Evaluations Course Evaluations http://www.siggraph.org/courses_evaluation 4 Random Individuals will win an ATI Radeon tm HD2900XT

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend