+ All Categories
Home > Documents > Local feature matching and multiple objects...

Local feature matching and multiple objects...

Date post: 04-Aug-2020
Category:
Upload: others
View: 37 times
Download: 0 times
Share this document with a friend
18
Introduction Introduction Which are the similar geometrical features between these images ?
Transcript
Page 1: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Introduction

Which are the similar geometrical features between these images ?

Page 2: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Introduction

Which are the similar geometrical features between these images ?

Page 3: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Introduction

Is there any similar geometrical features between these images ?

Page 4: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Introduction

Page 5: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Motivation and requirements

Problem: finding correspondences between images.

Applicationscomparison of images content;research in a database;object detection/recognition;image stiching;3D reconstruction, stereo...

Usual requirement: invariance or robustness toillumination (contrast changes),scale and viewpoint (local similarity, affine or projective transformations),occlusion,noise

Page 6: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Motivation and requirements

Page 7: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Methodology

Four steps1 Extraction of local features: invariance or robustness requirements;2 Feature comparison: distance between features;3 Decision: thresholds on the distance and matching;4 Grouping of previous matching in coherent rigid transformations.

A B

Page 8: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Methodology

Four steps1 Extraction of local features: invariance or robustness requirements;2 Feature comparison: distance between features;3 Decision: thresholds on the distance and matching;4 Grouping of previous matching in coherent rigid transformations.

A B

Page 9: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Methodology

Four steps1 Extraction of local features: invariance or robustness requirements;2 Feature comparison: distance between features;3 Decision: thresholds on the distance and matching;4 Grouping of previous matching in coherent rigid transformations.

A B

Page 10: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Introduction

Methodology

Four steps1 Extraction of local features: invariance or robustness requirements;2 Feature comparison: distance between features;3 Decision: thresholds on the distance and matching;4 Grouping of previous matching in coherent rigid transformations.

A B

Page 11: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Part I

Local descriptors

Page 12: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local representation choice

Many local features have been proposed in the literature:level lines pieces: [Lisani 2001], [Muse et al 2006];local descriptors: SIFT [Lowe, 1999], GLOH [Mikolajczyk, Schmid,2005], PCA-SIFT [Ke, Sukthankar, 2004], SURF [Bay et al, 2006], ShapeContext [Belongie, Malik, 2000] etc...region descriptors: shapes [Monasse, Guichard, 2000], MSER [Mataset al, 2002].

Page 13: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local feature extraction

1 Discrete image u2 Linear scale-space representation8�, u� = g� ⇤ u

3 Local extrema (~x , �) in space andscale of �2�u�

4 Harris criterion to eliminate edgepoints! interest points (~x , �).

5 Main orientations (direction ofru�) assigned at each point!interest points (~x , �, ✓).

Page 14: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local feature extraction

1 Discrete image u2 Linear scale-space representation8�, u� = g� ⇤ u

3 Local extrema (~x , �) in space andscale of �2�u�

4 Harris criterion to eliminate edgepoints! interest points (~x , �).

5 Main orientations (direction ofru�) assigned at each point!interest points (~x , �, ✓).

Page 15: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local feature extraction

1 Discrete image u2 Linear scale-space representation8�, u� = g� ⇤ u

3 Local extrema (~x , �) in space andscale of �2�u�

4 Harris criterion to eliminate edgepoints! interest points (~x , �).

5 Main orientations (direction ofru�) assigned at each point!interest points (~x , �, ✓).

Page 16: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local feature extraction

1 Discrete image u2 Linear scale-space representation8�, u� = g� ⇤ u

3 Local extrema (~x , �) in space andscale of �2�u�

4 Harris criterion to eliminate edgepoints! interest points (~x , �).

5 Main orientations (direction ofru�) assigned at each point!interest points (~x , �, ✓).

Page 17: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local feature extraction

1 Discrete image u2 Linear scale-space representation8�, u� = g� ⇤ u

3 Local extrema (~x , �) in space andscale of �2�u�

4 Harris criterion to eliminate edgepoints! interest points (~x , �).

5 Main orientations (direction ofru�) assigned at each point!interest points (~x , �, ✓).

Page 18: Local feature matching and multiple objects detectionhelios.mi.parisdescartes.fr/.../TP3/slides_SIFT.pdf · Local feature extraction SIFT Local feature extraction 1 Discrete image

Local feature extraction SIFT

Local feature representation: example of SIFT descriptors [Lowe, 1999]

Construction of a local descriptor a at each interest point (~x , �, ✓).

Mask (e.g. a square, a disk) around ~x :M sectors,size proportional to �.orientation given by ✓

Descriptor a = (a1, . . . aM)

am = normalized histogram of the gradientorientation (*), weighted by the gradientmagnitude, in the mth sector.(*) Orientations defined with respect to the referencedirection ✓.


Recommended