Contour Detection and Hierarchical Image Segmentation

Report
CONTOUR DETECTION AND
HIERARCHICAL IMAGE SEGMENTATION
P. Arbelaez, M. Maire, C. Fowlkes, J. Malik. Contour Detection and Hierarchical
image Segmentation. IEEE Trans. on PAMI , 2011.
Student: Hsin-Min Cheng
Advisor: Sheng-Jyh Wang
1
Outline
 Introduction
 Contour Detection
 Hierarchical Segmentation
 Results
 Conclusion
2
Introduction
 Contour
Original Image
Contour
3
Introduction
 Segmentation
Original Image
Segmentation
4
Introduction
 From Contour to Segmentation
Original Image
Contour
Segmentation
5
Introduction
 Goal
 Contour Detection
 Hierarchical Segmentation from Contours
Original Image
Contour
Segmentation
6
Outline
 Introduction
 Contour Detection
 Hierarchical Segmentation
 Results
 Conclusion
7
Contour Detection
1. Learn local boundary cues
2. Global framework to capture
closure, continuity
3. Local Cues and global cues
combination
8
Contour Detection
 Learn local boundary cues
Image
Local Boundary Cues
Brightness
Color
Cue Combination
Model
Texture
9
Contour Detection
 Learn local boundary cues
 Brightness
 L*a*b* colorspace
 Color
 L*a*b* colorspace
 Texture
 Convolve with 17 filters
Filters for creating textons
10
Contour Detection
 Learn local boundary cues
 Oriented gradient of histograms
 Example
 Gradient magnitude G at location(x, y)
 Three scales of r
ure[ ,  , 2 ]
2
11
Contour Detection
 Learn local boundary cues
 Local Cues Combination

r  [ ,  , 2 ]
ure
2
12
Contour Detection
 Global framework to capture closure, continuity
V: image pixels
E: connections between pairs of nearby pixels
=>Build a weighted graph G=(V,E) from image
13
Contour Detection
 Global framework to capture closure, continuity
14
Contour Detection
 Local Cues and global cues combination
Local Cues
Global cues
15
Outline
 Introduction
 Contour Detection
 Hierarchical Segmentation
 Results
 Conclusion
16
Hierarchical Segmentation
 Multiple Segmentations
 Fixed resolution
 Hierarchy of Segmentations
 Flexible resolution adjustment
17
Hierarchical Segmentation
1. From contours to segmentation
2. Hierarchical segmentation by
iterative merging
18
Hierarchical Segmentation
 From contours to segmentation
 Watershed Transform
 Concept
19
Hierarchical Segmentation
 From contours to segmentation
 Watershed Transform
 Example
20
Hierarchical Segmentation
 From contours to segmentation
 Watershed Transform
Artifacts
Boundary strength
Weight each arc
21
Hierarchical Segmentation
 From contours to segmentation
 Oriented Watershed Transform
OWT
WT
Hierarchical Segmentation
 Hierarchical segmentation by iterative
merging
 Hierarchical segmentation
 Example
23
Brief Summary
Original Image
- Local cues
- Global cues
Oriented Gradient
of histograms
Contour
Oriented Watershed Transform
Hierarchical Segmentation
Iterative Merging
24
Outline
 Introduction
 Contour Detection
 Hierarchical Segmentation
 Results
 Conclusion
25
Result
26
Result
27
Result
 BSDS300 Dataset
Evaluation of contour detector
Evaluation of segmentation algorithms
28
Outline
 Introduction
 Contour Detection
 Hierarchical Segmentation
 Results
 Conclusion
29
Conclusion
 A high performance contour detector, combining
local and global image information
 A method to transform any contour detector signal
into a hierarchy of regions while preserving
contour quality
30
Reference
 P. Arbelaez, M. Maire, C. Fowlkes and J. Malik. Contour Detection
and Hierarchical Image Segmentation. IEEE TPAMI, Vol. 33, No. 5,
pp. 898-916, May 2011
 P. Arbelaez, M. Maire, C. Fowlkes and J. Malik. From Contours to
Regions: An Empirical Evaluation. In CVPR 2009.
 P. Arbelaez and L. Cohen. Constrained Image Segmentation from
Hierarchical Boundaries. In CVPR 2008.
31

similar documents