0% found this document useful (0 votes)
2 views71 pages

4.MachineLearning_features_part2

The document provides an overview of local features in image processing, focusing on scale and affine invariant interest point detectors, including Harris, SIFT, and their evaluation. It discusses the importance of scale adaptation, multi-scale matching algorithms, and the performance of various region descriptors. The conclusion emphasizes the complementary nature of different detectors and the superior performance of SIFT-based descriptors in various conditions.

Uploaded by

Ari
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
2 views71 pages

4.MachineLearning_features_part2

The document provides an overview of local features in image processing, focusing on scale and affine invariant interest point detectors, including Harris, SIFT, and their evaluation. It discusses the importance of scale adaptation, multi-scale matching algorithms, and the performance of various region descriptors. The conclusion emphasizes the complementary nature of different detectors and the superior performance of SIFT-based descriptors in various conditions.

Uploaded by

Ari
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Overview

• Introduction to local features

• Harris interest points + SSD, ZNCC, SIFT

• Scale & affine invariant interest point detectors

• Evaluation and comparison of different detectors

• Region descriptors and their performance


Scale invariance - motivation

• Description regions have to be adapted to scale changes

• Interest points have to be repeatable for scale changes


Harris detector + scale changes

Repeatability rate

| {(a i , b i ) | dist ( H (a i ), b i )   } |
R ( ) 
max(| a i |, | b i |)
Scale adaptation
Scale change between two images

 x1   x2   sx1 
I1    I 2    I 2  
 y1   y2   sy1 

Scale adapted derivative calculation


Scale adaptation
Scale change between two images

 x1   x2   sx1 
I1    I 2    I 2  
 y1   y2   sy1 

Scale adapted derivative calculation

 x1   x2 
I1    Gi i ( )  s I 2    Gi i ( s )
nn

 y1   y2 
1 n 1 n
Scale adaptation
x ( ) Lx Ly ( )
~  L2

G ( )   
L L
 x y ( ) L 2
y ( ) 

where Li ( ) are the derivatives with Gaussian convolution


Scale adaptation
x ( ) Lx Ly ( )
~  L2

G ( )   
L L
 x y ( ) L 2
y ( ) 

where Li ( ) are the derivatives with Gaussian convolution

Scale adapted auto-correlation matrix

 x ( s ) Lx Ly ( s )
2
~ L
s G ( s )  
2

L L
 x y ( s  ) L 2
y ( s  ) 
Harris detector – adaptation to scale

R ( )  {(a i , b i ) | dist ( H (a i ), b i )   }
Multi-scale matching algorithm

s 1

s3

s5
Multi-scale matching algorithm

s 1
8 matches
Multi-scale matching algorithm

Robust estimation of a global


s 1
affine transformation
3 matches
Multi-scale matching algorithm

s 1
3 matches

s3
4 matches
Multi-scale matching algorithm

s 1
3 matches

s3
4 matches

highest number of matches

correct scale s5


16 matches
Matching results

Scale change of 5.7


Matching results

100% correct matches (13 matches)


Scale selection
• For a point compute a value (gradient, Laplacian etc.) at
several scales
• Normalization of the values with the scale factor
e.g. Laplacian | s 2 ( Lxx  Lyy ) |

• Select scale s  at the maximum → characteristic scale

| s 2 ( Lxx  Lyy ) |

scale

• Exp. results show that the Laplacian gives best results


Scale selection
• Scale invariance of the characteristic scale

s
norm. Lap.

scale
Scale selection
• Scale invariance of the characteristic scale

s
norm. Lap.

norm. Lap.

scale scale

 
• Relation between characteristic scales s  s1  s2
Scale-invariant detectors
• Harris-Laplace (Mikolajczyk & Schmid’01)

• Laplacian detector (Lindeberg’98)

• Difference of Gaussian (Lowe’99)

Harris-Laplace Laplacian
Harris-Laplace

multi-scale Harris points

selection of points at
maximum of Laplacian

invariant points + associated regions [Mikolajczyk & Schmid’01]


Matching results

213 / 190 detected interest points


Matching results

58 points are initially matched


Matching results

32 points are matched after verification – all correct


Matching results

all matches are correct (33)


Laplacian of Gaussian (LOG)

LOG  Gxx ( )  Gyy ( )


LOG detector
Detection of maxima and minima of Laplacian in scale space
Difference of Gaussian (DOG)
• Difference of Gaussian approximates the Laplacian
DOG  G (k )  G ( )
DOG detector
• Fast computation, scale space processed one octave at a
time
Local features - overview

• Scale invariant interest points

• Affine invariant interest points

• Evaluation of interest points

• Descriptors and their evaluation


Affine invariant regions - Motivation
• Scale invariance is not sufficient for large baseline changes

detected scale invariant region

projected regions, viewpoint changes can locally


be approximated by an affine transformation A
Affine invariant regions - Motivation
Affine invariant regions - Example
Harris/Hessian/Laplacian-Affine
• Initialize with scale-invariant Harris/Hessian/Laplacian
points

• Estimation of the affine neighbourhood with the second


moment matrix [Lindeberg’94]

• Apply affine neighbourhood estimation to the scale-


invariant interest points [Mikolajczyk & Schmid’02,
Schaffalitzky & Zisserman’02]

• Excellent results in a recent comparison


Affine invariant regions
• Based on the second moment matrix (Lindeberg’94)

 Lx (x, D ) Lx Ly (x, D )
2

M   (x, I , D )   D G( I )  
2

Lx Ly (x, D ) Ly (x, D ) 
2

• Normalization with eigenvalues/eigenvectors

x  M x 2
Affine invariant regions

x R  Ax L

 1
 1

xL  M xL
2
L xR  M xR 2
R

 
x R  Rx L

Isotropic neighborhoods related by image rotation


Affine invariant regions - Estimation
• Iterative estimation – initial points
Affine invariant regions - Estimation
• Iterative estimation – iteration #1
Affine invariant regions - Estimation
• Iterative estimation – iteration #2
Affine invariant regions - Estimation
• Iterative estimation – iteration #3, #4
Harris-Affine versus Harris-Laplace

Harris-Affine Harris-Laplace
Harris/Hessian-Affine

Harris-Affine

Hessian-Affine
Harris-Affine
Hessian-Affine
Matches

22 correct matches
Matches

33 correct matches
Maximally stable extremal regions (MSER) [Matas’02]

• Extremal regions: connected components in a thresholded


image (all pixels above/below a threshold)

• Maximally stable: minimal change of the component


(area) for a change of the threshold, i.e. region remains
stable for a change of threshold

• Excellent results in a recent comparison


Maximally stable extremal regions (MSER)

Examples of thresholded images

high threshold

low threshold
MSER
Overview
• Introduction to local features

• Harris interest points + SSD, ZNCC, SIFT

• Scale & affine invariant interest point detectors

• Evaluation and comparison of different detectors

• Region descriptors and their performance


Evaluation of interest points
• Quantitative evaluation of interest point/region detectors
– points / regions at the same relative location and area

• Repeatability rate : percentage of corresponding points

• Two points/regions are corresponding if


– location error small
– area intersection large

• [K. Mikolajczyk, T. Tuytelaars, C. Schmid, A. Zisserman, J. Matas, F.


Schaffalitzky, T. Kadir & L. Van Gool ’05]
Evaluation criterion

# corresponding regions
repeatability  100%
# detected regions
Evaluation criterion

# corresponding regions
repeatability  100%
# detected regions
intersection
overlap error  (1  ) 100%
union

2% 10% 20% 30% 40% 50% 60%


Dataset
• Different types of transformation
– Viewpoint change
– Scale change
– Image blur
– JPEG compression
– Light change

• Two scene types


– Structured
– Textured

• Transformations within the sequence (homographies)


– Independent estimation
Viewpoint change (0-60 degrees )

structured scene

textured scene
Zoom + rotation (zoom of 1-4)

structured scene

textured scene
Blur, compression, illumination

blur - structured scene blur - textured scene

light change - structured scene jpeg compression - structured scene


Comparison of affine invariant detectors
Viewpoint change - structured scene
repeatability % # correspondences

reference image 20 40 60
Comparison of affine invariant detectors
Scale change
repeatability % repeatability %

reference image 2.8 reference image 4


Conclusion - detectors
• Good performance for large viewpoint and scale changes

• Results depend on transformation and scene type, no one best


detector

• Detectors are complementary


– MSER adapted to structured scenes
– Harris and Hessian adapted to textured scenes

• Performance of the different scale invariant detectors is very similar


(Harris-Laplace, Hessian-Laplace, LoG and DOG)

• Scale-invariant detector sufficient up to 40 degrees of viewpoint


change
Overview
• Introduction to local features

• Harris interest points + SSD, ZNCC, SIFT

• Scale & affine invariant interest point detectors

• Evaluation and comparison of different detectors

• Region descriptors and their performance


Region descriptors

• Normalized regions are


– invariant to geometric transformations except rotation
– not invariant to photometric transformations
Descriptors

• Regions invariant to geometric transformations except


rotation
– rotation invariant descriptors
– normalization with dominant gradient direction

• Regions not invariant to photometric transformations


– invariance to affine photometric transformations
– normalization with mean and standard deviation of the image patch
Descriptors
• Sampled image patch
– descriptor dimension is 81

 

• Gaussian derivative-based descriptors


– Differential invariants (Koenderink and van Doorn’87) (dim. 8)

*
Descriptors
• Gaussian derivative-based descriptors
– Steerable filters (Freeman and Adelson’91)

– “Steering the derivatives in the direction of an angle “

f ' ( )  I x cos  I y sin 


f ' ' ( )  I xx cos 2   2 IxIy sin  cos   I yy sin 2 

n, i  g  i /(n  1)  i  0...n

• Dominant gradient direction is rotation invariant


Descriptors
• SIFT [Lowe’99]
– 8 orientations of the gradient (dim. 128)
– 4x4 spatial grid
– normalization of the descriptor to norm one

3D histogram
image patch gradient x

 
y
Descriptors

• Moment invariants [Van Gool et al.’96]

• Shape context [Belongie et al.’02]

• SIFT with PCA dimensionality reduction

• Gradient PCA [Ke and Sukthankar’04]


Comparison criterion
• Descriptors should be
– Distinctive
– Robust to changes on viewing conditions as well as to errors of
the detector

• Detection rate (recall) 1


– #correct matches / #correspondences
• False positive rate
– #false matches / #all matches
• Variation of the distance threshold
– distance (d1, d2) < threshold
1

[K. Mikolajczyk & C. Schmid, PAMI’05]


Viewpoint change (60 degrees)
* *
Scale change (factor 2.8)
* *
Conclusion - descriptors
• SIFT based descriptors perform best

• Significant difference between SIFT and low dimension


descriptors as well as cross-correlation

• Robust region descriptors better than point-wise


descriptors

• Performance of the descriptor is relatively independent of


the detector
Available on the internet

[Link]

• Binaries for detectors and descriptors


– Building blocks for recognition systems

• Carefully designed test setup


– Dataset with transformations
– Evaluation code in matlab
– Benchmark for new detectors and descriptors

You might also like