Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer...
-
Upload
patricia-marilynn-davidson -
Category
Documents
-
view
224 -
download
3
Transcript of Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer...
![Page 1: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/1.jpg)
http://astro.temple.edu/~siddu1
Image Segmentation Using Region Growing and Shrinking
Computer Graphics and Image Processing
Professor : Dr. Longin Jan Latecki
![Page 2: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/2.jpg)
http://astro.temple.edu/~siddu2
Contents
Brief introduction to Image segmentation Types of Image segmentation Region growing and Shrinking (split /merge )
method Applications of Image segmentation Results
![Page 3: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/3.jpg)
http://astro.temple.edu/~siddu3
Introduction
The shape of an object can be described in terms of:
Its boundary – requires image edge detection The region it occupies – requires image
segmentation in homogeneous regions, Image regions generally have homogeneous characteristics (e.g. intensity, texture)
![Page 4: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/4.jpg)
http://astro.temple.edu/~siddu4
Introduction- cont.d
The goal of Image Segmentation is to find regions that represent objects or meaningful parts of objects. Major problems of image segmentation are result of noise in the image.
An image domain X must be segmented in N different regions R(1),…,R(N)
The segmentation rule is a logical predicate of the form P(R)
![Page 5: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/5.jpg)
http://astro.temple.edu/~siddu5
Introduction- cont.d
Image segmentation partitions the set X into the subsets R(i), i=1,…,N having the following propertiesX = i=1,..N U R(i)
R(i) ∩ R(j) = 0 for I ≠ j
P(R(i)) = TRUE for i = 1,2,…,N
P(R(i) U R(j)) = FALSE for i ≠ j
![Page 6: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/6.jpg)
http://astro.temple.edu/~siddu6
Introduction- cont.d
The segmentation result is a logical predicate of the form P(R,x,t)
x is a feature vector associated with an image pixel
t is a set of parameters (usually thresholds) A simple segmentation rule has the form:
P(R) : I(r,c) < T
![Page 7: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/7.jpg)
http://astro.temple.edu/~siddu7
Introduction- cont.d
In the case of color images the feature vector x can be three RGB image components {IR(r,c),IG(r,c),IB(r,c)
A simple segmentation rule may have the form:
P(R,x,t) : (IR(r,c) <T(R)) && (IG(r,c)<T(G))&&
(IB(r,c) < T(B))
![Page 8: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/8.jpg)
http://astro.temple.edu/~siddu8
Introduction- cont.d
A region is called connected if : ……. A pixel (x,y) is said to be adjacent to the pixel
(a,b) if it belongs to its immediate neighbourhood
The 4-neighbourhood of a pixel (x,y) is the set that includes its …..
The 8-neighbourhood of (x,y) is a superset of the 4-neighbourhood and contains the …..
![Page 9: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/9.jpg)
http://astro.temple.edu/~siddu9
Types
By Histogram ThresholdingBy Region Growing and ShrinkingBy Clustering in the color space
![Page 10: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/10.jpg)
http://astro.temple.edu/~siddu10
Region Growing
A simple approach to image segmentation is to start from some pixels (seeds) representing distinct image regions and to grow them, until they cover the entire image
For region growing we need a rule describing a growth mechanism and a rule checking the homogeneity of the regions after each growth step
![Page 11: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/11.jpg)
http://astro.temple.edu/~siddu11
Region Growing – cont.d
The growth mechanism – at each stage k and for each region Ri(k), i = 1,…,N, we check if there are unclassified pixels in the 8-neighbourhood of each pixel of the region border
Before assigning such a pixel x to a region Ri(k),we check if the region homogeneity:
P(Ri(k) U {x}) = TRUE , is valid
![Page 12: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/12.jpg)
http://astro.temple.edu/~siddu12
Region Growing – cont.d
The arithmetic mean m and standard deviation sd of a class Ri having n pixels:
M = (1/n)(r,c)€R(i) ∑ I(r,c)s.d = Square root((1/n)(r,c)€R(i) ∑[I(r,c)-
M]2)Can be used to decide if the merging of the two regions R1,R2 is allowed, if|M1 – M2| < (k)s.d(i) , i = 1, 2 , two regions are merged
![Page 13: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/13.jpg)
http://astro.temple.edu/~siddu13
Region Growing – cont.d
Homogeneity test: if the pixel intensity is close to the region mean value
|I(r,c) – M(i)| <= T(i) Threshold Ti varies depending on the region
Rn and the intensity of the pixel I(r,c).It can be chosen this way:
T(i) = { 1 – [s.d(i)/M(i)] } T
![Page 14: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/14.jpg)
http://astro.temple.edu/~siddu14
Split / Merge
The opposite approach to region growing is region shrinking ( splitting ).
It is a top-down approach and it starts with the assumption that the entire image is homogeneous
If this is not true , the image is split into four sub images
This splitting procedure is repeated recursively until we split the image into homogeneous regions
![Page 15: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/15.jpg)
http://astro.temple.edu/~siddu15
Split / Merge
If the original image is square N x N, having dimensions that are powers of 2(N = 2n):
All regions produced but the splitting algorithm are squares having dimensions M x M , where M is a power of 2 as well (M=2m,M<= n).
Since the procedure is recursive, it produces an image representation that can be described by a tree whose nodes have four sons each
Such a tree is called a Quadtree.
![Page 16: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/16.jpg)
http://astro.temple.edu/~siddu16
Split / Merge
Quadtree
R0 R1
R2R3
R0
R1
R00 R01 R02 R04
![Page 17: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/17.jpg)
http://astro.temple.edu/~siddu17
Split / Merge
Splitting techniques disadvantage, they create regions that may be adjacent and homogeneous, but not merged.
Split and Merge method – It is an iterative algorithm that includes both splitting and merging at each iteration:
![Page 18: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/18.jpg)
http://astro.temple.edu/~siddu18
Split / Merge
If a region R is inhomogeneous (P(R)= False) then is split into four sub regions
If two adjacent regions Ri,Rj are homogeneous (P(Ri U Rj) = TRUE), they are merged
The algorithm stops when no further splitting or merging is possible
![Page 19: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/19.jpg)
http://astro.temple.edu/~siddu19
Split / Merge
The split and merge algorithm produces more compact regions than the pure splitting algorithm
![Page 20: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/20.jpg)
http://astro.temple.edu/~siddu20
Applications
3D – Imaging : A basic task in 3-D image processing is the segmentation of an image which classifies voxels/pixels into objects or groups. 3-D image segmentation makes it possible to create 3-D rendering for multiple objects and perform quantitative analysis for the size, density and other parameters of detected objects.
Several applications in the field of Medicine like magnetic resonance imaging (MRI).
![Page 21: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/21.jpg)
http://astro.temple.edu/~siddu21
Results – Region grow
![Page 22: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/22.jpg)
http://astro.temple.edu/~siddu22
Results – Region Split
![Page 23: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/23.jpg)
http://astro.temple.edu/~siddu23
Results – Region Split and Merge
![Page 24: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/24.jpg)
http://astro.temple.edu/~siddu24
Results – Region growing
![Page 25: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/25.jpg)
http://astro.temple.edu/~siddu25
Results – Region Split
![Page 26: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/26.jpg)
http://astro.temple.edu/~siddu26
Results – Region Split and Merge
![Page 27: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/27.jpg)
http://astro.temple.edu/~siddu27
References
“Digital Image Processing” – Algorithms and Application , A multimedia approach.
Prof. Ioannis Pitas Computer Vision and Image Processing: A
Practical Approach CVIP Tools software
![Page 28: Http://astro.temple.edu/~siddu 1 Image Segmentation Using Region Growing and Shrinking Computer Graphics and Image Processing Professor : Dr. Longin Jan.](https://reader035.fdocuments.us/reader035/viewer/2022062313/56649d705503460f94a522ba/html5/thumbnails/28.jpg)
http://astro.temple.edu/~siddu28
Thank You