Computer Vision: Filtering Operations (Do, 23.10.2014)

Anmeldung erforderlich

RWTH

Für RWTH-Angehörige und aus dem RWTH-Netz verfügbar

Anmelden
  • Einbetten

Kapitel:

00:00:00
Lecture 3: Linear Filters
00:01:42
Demo "Haribo Classification"
00:03:15
You Can Do It At Home...
00:04:48
Course Outline
00:05:29
Motivation
00:05:55
Topics of This Lecture
00:07:03
Common Types of Noise
00:11:50
Gaussian Noise
00:12:50
First Attempt at a Solution
00:13:59
Moving Average in 2D
00:15:12
Correlation Filtering
00:18:25
Convolution
00:20:06
Correlation vs. Convolution
00:21:46
Shift Invariant Linear System
00:23:27
Properties of Convolution
00:27:44
Averaging Filter
00:28:57
Smoothing by Averaging
00:31:42
Smoothing with a Gaussian
00:33:04
Gaussian Smoothing
00:35:30
Gaussian Smoothing: variance
00:36:39
Gaussian Smoothing: kernel size
00:39:49
Gaussian Smoothing in Matlab
00:41:22
Effect of Smoothing
00:42:56
Efficient Implementation
00:45:33
Filtering: Boundary Issues
00:47:11
Filtering: Boundary Issues: Methods
00:52:44
Filtering: Boundary Issues: Methods (MATLAB)
00:53:48
Topics of This Lecture
00:54:21
Why Does This Work?
00:57:21
The Fourier Transform in Cartoons
00:58:22
Fourier Transforms of Important Functions
01:02:45
Duality
01:04:27
Effect of Convolution
01:06:38
Effect of Filtering
01:11:06
Low-Pass vs. High-Pass
01:14:22
Quiz: What Effect Does This Filter Have?
01:15:21
Sharpening Filter
01:16:51
Application: High Frequency Emphasis
01:18:56
Topics of This Lecture
01:19:11
Non-Linear Filters: Median Filter
01:20:26
Median Filter
01:20:53
Median Filter: edge preserving
01:21:46
Median vs. Gaussian Filtering
01:24:33
Topics of This Lecture
01:24:49
Motivation: Fast Search Across Scales
01:25:44
Image Pyramid
01:26:03
How Should We Go About Sampling?
01:28:11
Fourier Interpretation: Discrete Sampling
01:29:58
Sampling and Aliasing
01:31:20
Sampling and Aliasing: Nyquist theorem
01:32:51
Aliasing in Graphics
01:33:09
Resampling with Prior Smoothing
01:34:37
References and Further Reading