parlab teaching

MACHINE VISION

Educational program

Swiss Innovation Academy / CSEM Brazil - Trainee Programme

1 day, morning 4 periods, afternoon 3 periods.

Content

Chap 1: Introduction  2

Machine vision (MV). 3

Vision    4

MV applications. 5

Chap 2: Image acquisition and image properties. 2

Image acquisition. 2

Image sources. 3

Image properties. 8

Image  9

Camera image. 10

Imaging geometry. 11

Image sampling. 12

Beyond Nyquist 13

Image quantizing. 14

Information. 15

Image forming. 17

Signal distribution. 18

Histogram examples. 19

Signal to noise ratio. 20

Chap 3: Image correction and filtering. 2

Geometric corrections. 2

Geometric correction problem.. 3

Geometric correction procedure. 4

Geometric transform.. 5

Specific transformations. 6

Point operations. 8

Gray level mapping. 9

Changing contrast 10

Normalisation to maximum range. 11

Color mapping. 12

g(f) gamma. 13

Image math. 14

Example of image subtraction. 15

Spatial filtering. 16

Linear invariant systems. 17

Convolution with a two-dimensional (2D) kernel 19

Low pass or smoothing filters. 20

Separable kernel 21

High pass filters. 22

Exemples  24

Non-linear filtering. 25

Linear versus median filtering. 26

Chap 4: image matching. 2

Comparing images. 3

Template maching problem.. 4

Template maching procedure. 5

Similarity and dissimilarity measures. 6

Properties of similarity/dissimilarity measures. 7

Application. 8

Chap 5: thresholding and segmentation. 3

Thresholding. 3

Thresholding. 4

Examples  6

Example of thresholding a multidimensional color image. 7

Segmentation. 8

Neighborhood. 9

Path C   10

Region and connex region. 11

Segmentation and partition. 12

Homogeneity. 13

Contour and region. 14

Blob coloring. 15

Example  17

Segmentation of gray-level images. 18

Graph based region growing. 19

Split and merge. 20

Edge detection. 21

Edge of one dimensional f(x) 22

Edge of f(x,y) 23

Differential gradient (DG) approach. 24

Example  25

Template matching (TM) approach. 26

Optimal edge detection. 27

Hysteresis thresholding. 28

Exemple  29

Laplacian approach. 30

Laplacien of smooth image. 31

Example  33

Contour modeling: snakes, marching cubes & level set methods  34

Chap 6: Mathematical morphology. 3

Binary morphology. 3

Binary operations. 4

Transformation by a structuring element 5

Erosion  6

Dilatation  7

Erosion and dilatation examples. 8

Duality of erosion and dilation. 9

Illustration  10

Iterative erosion and dilatation. 11

Opening  12

Closing  13

Duality of opening and closing. 14

Properties of erosion, dilation, opening and closing. 15

Examples  16

Applications of opening and closing. 17

Application: contour detection. 18

Opening as a size filter 19

Conditional and geodesic dilation. 20

Geodesic opening as size filter 22

Iterative size filter 23

Classes of structuring elements. 24

Example anisotropic filtering. 25

Gray level morphology. 27

Erosion and dilation. 28

Examples  29

Opening and closing. 31

Rank Order Based Filters ROBF. 32

Chap 7: features. 2

Contour features. 2

Contour following. 3

Contour representation. 5

Contour as a continuous parametric function. 6

Contour as a discrete parametric function. 7

Curvilinear (s) versus polar (x) features. 9

Problems with the centroid feature approach. 11

Comparison. 12

Shape features. 13

Simple features. 14

Application. 18

Moments of a binary shape. 19

Equivalent ellipse. 21

Signatures  22

Features for gray-level images. 23

First order features. 24

Second order features. 25

Texture descriptors. 26

Application. 27

Particle shape analysis. 28

Representation in feature space. 30

More features. 32

On-line learning tools

HIPR2
 http://www.dai.ed.ac.uk/HIPR2
is a collection of free web-based tutorial materials on the 50 most commonly used image processing operators. Each operator has an individual JAVA exercise program, plus there is a JAVA-based Khoros-like drag and drop workspace tableau. It works under Internet Explorer and Mozilla as well as Netscape (Java 1.2 or higher is needed). HIPR2 just had its 65,000th user.

CVonline
http://www.dai.ed.ac.uk/CVonline/
is a free web-based compendium of topics in computer vision. CVonline just had its 100,000th visitor, each of which accesses about 4 topics. There are 1319 topics of which 793 have content so far.

Lectures on-line

U. Central Florida,Dr. Ylmaz
Lappeenranta University of Technology (LUT), Dptmt IT

Références

Recommended book
E.R. Davis - "Machine Vision - Theory, Algorithms and Practice" (Elsevier, 3rd Edition, 2005)

Books on-line
http://homepages.inf.ed.ac.uk/rbf/CVonline/books.htm

Image Processing Fundamentals
http://www.ph.tn.tudelft.nl/Courses/FIP/noframes/fip.html

The computer vision homepage
http://www-2.cs.cmu.edu/afs/cs/project/cil/ftp/html/vision.html


hu / 30.05.2013