Presentation is loading. Please wait.

Presentation is loading. Please wait.

Introduction Computational Photography Seminar: EECS 395/495

Similar presentations


Presentation on theme: "Introduction Computational Photography Seminar: EECS 395/495"— Presentation transcript:

1 Introduction Computational Photography Seminar: EECS 395/495
Northwestern University

2 About Me New faculty in EECS Research Teaching
Joined in September 2012 Ph.D. 2011, from Columbia University PostDoc , Columbia University Research Computational Photography/Imaging, Optics, Computer Vision, Computer Graphics Teaching Computational Photography Seminar Introduction to Computational Photography Introduction to Programming in Python

3 Previous Research 3D Displays: Perspecta at Actuality Systems
Front View of 3D Graphic on Perspecta Side View of 3D Graphic on Perspecta Perspecta 3D Display

4 Previous Research Computer Graphics: Light Field Transfer

5 Traditional F/18 (Normalized)
Previous Research Computational Photography: Extending DOF Traditional F/1.8 EDOF Image F/1.8 (Captured) Traditional F/18 (Normalized) EDOF Image F/1.8 (Deblurred)

6 Previous Research Computational Vision: Depth from Diffusion

7 What About You? Who are you? What are you interested in?
What department are you in? Are you an undergraduate or graduate student? Why do you want to take this course? What do you hope to get out of it?

8 What is Computer Vision?
Camera Lighting Vision System Scene Scene Description

9 Computational Photography
Novel Illumination Generalized Optics Light Sources Modulators 4D Incident Lighting Computational Cameras Generalized Sensor Generalized Optics Processing Novel Displays Recreate 4D Lightfield Scene: 8D Ray Modulator 4D Light Field 4 blocks : light, optics, sensors, processing, (display: light sensitive display) Slide Credits: Nayar and Rashkar

10 Course Overview This course IS about
Research in computational photography Reading, presenting and discussing research Final projects

11 Class Structure Monday 2-5pm Tech F280
Class will consist of minute presentations. Students will each give 2-3 in class presentations.

12 Course Webpage

13 Schedule Week 1: Introduction to Computational Photography
Week 2: Extended Depth of Field and Coded Aperture Week 3: Motion Deblurring and Video Processing Week 4: Light Field Imaging Week 5: Spectral and Multimodal Imaging Week 6: Light Transport Acquisition and Processing Week 7: Structured Illumination and 3D Capture Week 8: Computational Displays Week 9: Transient and Compressive Imaging Week 10: Final Projects

14 Week 2: Defocus Processing
Traditional F/1.8 EDOF Image F/1.8 (Deblurred) Traditional F/18 (Normalized) Topics: Extended Depth of Field (EDOF) Coded aperture imaging Coded aperture depth-from-defocus 3D Shape from a single image

15 Week 3: Motion Processing
Captured Image Flutter Shutter Camera Other Topics: Motion Invariant Photography Dark Flash Photography Video Magnification Deblurred Image

16 Week 4: Light Fields Lytro Plenoptic Stanford Light Field
Camera Array Lytro Plenoptic Light Field Camera Topics: Light Field Microscopy Light Field Superresolution Exotic Light Field Camera Designs

17 Week 5: Spectral Imaging
Topics: Statistics of Multispectral Images Multispectral Video Camera Multispectral Imaging and Relighting

18 Week 6: Light Transport Direct/Global Separation Direct Global Topics:
Helmholtz Stereo Dual Photography Inverse Light Transport Synthetic Aperture Imaging

19 Week 7: 3D Capture Topics: Outdoor structured light
Structured light and global illumination Motion aware structured light Gelsight – high resolution depth estimation

20 Week 8: Computational Displays
Lighting Sensitive Display Topics: Light Field Probes for Measuring 3D Layered Tomographic 3D Displays Compressive Parallax Barrier 3D Displays Light Field Displays for Optometry

21 Week 9: Exotic Cameras Transient/Femto-second Imaging Topics:
Recovering 3D shape around a corner Compressive Light Field Cameras Compressive Video Cameras Compressive Hyperspectral Cameras

22 Presentations Give a clear, concise explanation of your paper. Come to class prepared! See me beforehand if you need help with the paper Presentations should be 20 min total Motivation and related work (5 min) Intuition and technical idea (5 min) Results (5 min) Discussion (5 min) Submit a paper review form (see website) after presenting

23 Final Projects You can propose any final project as long as it is related to cameras, optics, lighting, image processing Proposals due Monday April 27 Final projects presented in class Monday June 1 Small budget for expenses Alternative (Literature Survey): Read 3 extra papers from the webpage and write a 6-8 page report summarizing the work. Report is also due June 4.

24 Paper Topics Let me know if you would like to suggest a different paper for your presentation or literature review

25 Grading Presentations and Discussions – 50% Paper Reviews - 10%
Final Project – 40% No Tests, Midterm, or Final

26 Final Projects Ideas Build your own 3D display
Build your own 3D display Use an LCD monitor or cellphone display Attach a lens array or printed transparency Calibrate the screen Render 3D images Interlace 3D images and load onto the display

27 Final Projects Ideas Build a light field probe
Build a light field probe Attach a color transparency to a lens array Visualize 3D surface normals for transparent objects Calculate surface shape from normals

28 Final Projects Ideas Process Lytro images:
Process Lytro images: Extract raw plenoptic data from a Lytro camera Calibrate the camera Compute the light field from the captured data Compute synthetic refocused images

29 Final Projects Ideas Structured Light 3D Scanning
Structured Light 3D Scanning Captured images 3D model Project sequences of binary patterns and capture images Use triangulation to calculate a 3D model Render the 3D model to visualize the results

30 Final Projects Ideas Direct and global separation
Direct and global separation Project sequences of binary patterns and capture images Simple Matlab code separates direct and global components

31 Final Projects Ideas More ideas at: Or
Or Propose your own project. It could be anything that relates to photography, cameras, or image processing.

32 Previous Projects

33

34

35 Previous Projects

36 Previous Projects

37 Previous Projects

38 Previous Projects

39 Previous Projects

40 Previous Projects

41 Previous Projects


Download ppt "Introduction Computational Photography Seminar: EECS 395/495"

Similar presentations


Ads by Google