31
LING 111 Teaching Demo By Tenghui Zhu ( [email protected] ) Introduction to Edge Detection Image Segmentation

LING 111 Teaching Demo By Tenghui Zhu ([email protected])[email protected] Introduction to Edge Detection Image Segmentation

Embed Size (px)

Citation preview

LING 111 Teaching DemoLING 111 Teaching Demo

By Tenghui Zhu ([email protected])

Introduction to Edge Detection

Image Segmentation

OverviewOverview

• Review on Image Data Structures

• Purpose of Image Segmentation

• Basic Image Segmentation Algorithm – Edge Detection

OverviewOverview

• Review on Image Data Structures

• Purpose of Image Segmentation

• Basic Image Segmentation Algorithm – Edge Detection

Image Data StructuresImage Data Structures

grayscale images : (Center)each pixel = a value between 0 - 255.

monochrome images: (Left)each pixel = a single 0 or 1 value

Color Images: (Right)each pixel = three values (red, green and blue - RGB).

Image Data StructuresImage Data Structures

grayscale images : (Center)each pixel = a value between 0 - 255.

monochrome images: (Left)each pixel = a single 0 or 1 value

Color Images: (Right)each pixel = three values (red, green and blue - RGB).

OverviewOverview

• Review on Image Data Structures

• Purpose of Image Segmentation

• Basic Image Segmentation Algorithm – Edge Detection

One Image Segmentation applicationOne Image Segmentation applicationOne Image Segmentation applicationOne Image Segmentation application

Any other Image Segmentation application ?Any other Image Segmentation application ?Any other Image Segmentation application ?Any other Image Segmentation application ?

• Satellite Images for CIA• Medical Images for NIH • Face Recognize for FBI • Etc…

OverviewOverview

• Review on Image Data Structures

• Purpose of Image Segmentation

• Basic Image Segmentation Algorithm – Edge Detection

Edge Detection—3x3 FilteringEdge Detection—3x3 Filtering

8 9 9 11 11 11 8 3 9 3 9 11 2 8 3 5 5 9 8 6 8 6 8 11 2 2 4 4 6 9 6 9 11 4 4 6

1/9 1/9 1/9

1/9 1/9 1/9

1/9 1/9 1/9

(3 + 5 + 5 + 8 + 6 + 8 + 4 + 4 + 6) / 9 = 5.44

Round(3 + 5 + 5 + 8 + 6 + 8 + 4 + 4 + 6) / 9 = 5or

Edge Detection—3x3 FilteringEdge Detection—3x3 Filtering

8 9 9 11 11 11 8 3 9 3 9 11 2 8 3 5 5 9 8 6 8 6 8 11 2 2 4 4 6 9 6 9 11 4 4 6

1/9 1/9 1/9

1/9 1/9 1/9

1/9 1/9 1/9

(3 + 5 + 5 + 8 + 6 + 8 + 4 + 4 + 6) / 9 = 5.44

Round(3 + 5 + 5 + 8 + 6 + 8 + 4 + 4 + 6) / 9 = 5or

55

Edge Detection FiltersEdge Detection Filters

R1 R2 R3 R4

Edge DetectionEdge Detection

Edge DetectionEdge Detection

Constrains of Edge Detection

OverviewOverview

• Review on Image Data Structures

• Purpose of Image Segmentation

• Basic Image Segmentation Algorithm – Edge Detection

Next time, we will talk about another basic Image Segmentation Algorithm -- Histogram

Thank you !Thank you !

Thank you !Thank you !

Thank you !Thank you !

Gray Scale Image - bimodalGray Scale Image - bimodal

Image of a Finger Print with light background

Bimodal - HistogramBimodal - Histogram

Image Histogram of finger print

Segmented ImageSegmented Image

Image after Segmentation

Gray Scale Image - MultimodalGray Scale Image - Multimodal

Original Image of lena

Multimodal HistogramMultimodal Histogram

Histogram of lena

Segmented ImageSegmented Image

Image after segmentation – we get a outline of her face, hat, shadow etc

Colour Image - MultimodalColour Image - Multimodal

Colour Image having multi-modal histogram

HistogramHistogram

Image Histogram for the three colour spaces

Segmented ImageSegmented Image

OverviewOverview

• Review on Image Segmentation– Algorithm (Without User Interaction)

• Edge Detection• Statistics (Histogram etc…)

– Tools (With User Interaction)• Magic Wand • Intelligent Scissors • Graph Cut

ConclusionsConclusions ConclusionsConclusions

GrabCut – powerful interactive extraction tool

Iterated Graph Cut based on colour and contrast

Regularized alpha matting by Dynamic Programming