magic camera master’s project defense by adam meadows project committee: dr. eamonn keogh dr. doug...

37
Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug

Upload: harriet-kerry-fitzgerald

Post on 18-Jan-2016

221 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Magic Camera

Master’s Project Defense

By

Adam Meadows

Project Committee:Dr. Eamonn KeoghDr. Doug Tolbert

Page 2: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Roadmap

• Problem

• Motivation

• Background

• Stepping Through Magic Camera

• Results

• Conclusion

• Future Work

Page 3: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Problem

• To organize an image containing a collection of objects in front of a solid background

Page 4: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Motivation

• Incorporation into Digital Cameras– Sorting Tables– Insect Boards

Page 5: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Background

• Multidimensional Scaling (MDS)– Transforms a dissimilarity matrix into a

collection of points in 2d (or 3d) space– Euclidean distances between the points

reflect the given dissimilarity matrix– Similar objects are spaced close together,

dissimilar objects are spaced farther apart

Page 6: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Stepping Through Magic Camera

• Identifying Objects

• Calculating Similarities

• Creating Resulting Image

Page 7: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Identifying Objects

• Convert to black and white image– Threshold: calculated automatically or specified

• Each connected comp treated as an object• Each obj. cropped by B-box + 5 pixel border• Edges of adjacent objects filtered out• Objects rotated to “face” same direction

Page 8: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Filtering Adjacent Objects

Page 9: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Object Rotation

• Find major axis– Align with image’s major axis

• Find centroid– Rotate so centroid is at bottom/left of obj

http://www.mathworks.com/access/helpdesk/help/toolbox/images/regionprops.html

Page 10: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Calculating Similarities

• Numerical representation of objects– Shape, color, texture

• Create dissimilarity matrix– Euclidean dist between each pair of objs

Page 11: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Shape

• Each object translated into a time series

• Dist from the center of obj to perimeter – Code provided by Dr. Keogh

Page 12: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Shape II

Page 13: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Color

• RGB values independently averaged– 1000 random pixels chosen– Pixels not unique (if obj < 1000 pixels)

Page 14: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Texture

• Std deviation of 9 pixel neighborhood – averaged over 1,000 random pixels– Pixels not unique (if obj < 1,000 pixels)

Page 15: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Creating New Image

• Extracting Background

• Finding New Positions

• Fixing Overlaps

Page 16: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Extracting Background

• Use B&W image to id background

• Independently avg RGB values

• Create a new solid background image– same dimensions as original image

Page 17: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Finding New Positions

• Use MDS to get coordinates for objs– Using dissimilarity matrix

• Reverse Y values– Images are indexed top-down

Page 18: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Fixing Overlaps

• Start placing objects in given order– Randomly chosen if not specified

• If overlap detected– Move object min dist to rectify– In one direction (up, down, left, right)

Page 19: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Fixing Overlaps II

Not

Page 20: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Results

Page 21: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 22: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 23: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Explanation

Page 24: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Explanation II

Page 25: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 26: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 27: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 28: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 29: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 30: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 31: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 32: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 33: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert
Page 34: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Conclusion

• Input image– Collection of objects on solid background

• Output image– Similar objects grouped close to each other– All objects “face” same direction

Page 35: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Future Work

• Develop color method– Try it with some real data (butterflies, etc.)

• Add combination of similarity measures– Shape & color, color & texture, etc.

• Add optional How-To– Display original image– User clicks an object– Line drawn to new location

Page 36: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Questions ?

Page 37: Magic Camera Master’s Project Defense By Adam Meadows Project Committee: Dr. Eamonn Keogh Dr. Doug Tolbert

Resources

• Slides available– http://www.cs.ucr.edu/~ameadows/msproject/slides.ppt

– http://www.cs.ucr.edu/~ameadows/msproject/slides-handout.pdf

• Report available– http://www.cs.ucr.edu/~ameadows/msproject/report.pdf

• Code available– http://www.cs.ucr.edu/~ameadows/msproject/magic_camera.zip