Computer Vision: Recognition with Local Features (Di, 16.12.2014)

Anmeldung erforderlich

RWTH

Für RWTH-Angehörige und aus dem RWTH-Netz verfügbar

Anmelden
  • Einbetten

Beschreibung:

Indexing with Local Features, Inverted file index, Visual vocabularies

Kapitel:

00:00:00
Lecture 14: Recognition with Local Features
00:00:34
A Script...
00:01:44
Course Outline
00:02:47
Recap: Local Feature Matching Outline
00:04:17
Recap: Automatic Scale Selection
00:05:52
Recap: Laplacian-of-Gaussian (LoG)
00:08:10
Recap: LoG Detector Responses
00:08:41
Recap: Key point localization with DoG
00:09:49
Recap: Harris-Laplace [Mikolajczyk '01]
00:10:42
Topics of This Lecture
00:11:58
Local Descriptors
00:14:07
Feature Descriptors
00:16:09
Feature Descriptors: SIFT
00:18:55
Overview: SIFT
00:21:12
Working with SIFT Descriptors
00:22:49
Local Descriptors: SURF
00:29:39
You Can Try It At Home...
00:31:15
Topics of This Lecture
00:31:23
Applications of Local Invariant Features
00:31:56
Wide-Baseline Stereo
00:33:11
Automatic Mosaicing
00:33:28
Panorama Stitching
00:33:52
Recognition of Specific Objects, Scenes
00:34:18
Recognition of Categories
00:34:31
Value of Local Features
00:36:41
Topics of This Lecture
00:36:45
Recognition with Local Features
00:38:44
Concepts: Warping vs. Alignment
00:41:31
Parametric (Global) Warping
00:43:44
What Can be Represented by a 2×2 Matrix?
00:47:38
2D Linear Transforms
00:48:17
Homogeneous Coordinates
00:49:14
Basic 2D Transformations
00:51:36
2D Affine Transformations
00:52:20
Projective Transformations
00:55:52
Alignment Problem
01:01:03
Let's Start with Affine Transformations
01:01:48
Fitting an Affine Transformation
01:03:44
Recall: Least Squares Estimation
01:09:04
Fitting an Affine Transformation
01:11:43
Homography
01:15:03
Fitting a Homography
01:24:41
Image Warping with Homographies
01:26:27
Uses: Analyzing Patterns and Shapes
01:26:32
References and Further Reading