The course is given in Period 2  (~Nov-Dec),  in the Swedish Academic year.

Examiner:

J. Bigun                  josef.bigun@ide.hh.se

Laboratory Exercises:

M. Persson         martin.persson@ide.hh.se    (Further information: Laboratory Exercises)

Litterature:

[1] B. Jahne, "Digital image processing", 4'th edition, Springer, (1997).

(The book comes with a CD-rom which includes exercises as well as the software to run these exercises with, on a windows NT or windows 95/98 environment.)

[2] Material and exercises  distributed or pointed to at the lectures
 
 

Examination:

The examination which results in a grade consists in achieving one or more of the following:

• Participation and completion of  2/3 of the laboratory exercises. (=  Grade 3)
• Suggestion and oral presentation of theory.  (+ 1 Grade)
•  Implementation and demonstration  of two exercises.  (+ 1 Grade).

Contents (Numbers refer to chapters in Jahne, 1997)

 
Signal representation Fourier transform (FT) of continuous functions (To be distributed)

Properties of FT: Hermiticity, Convolution (To be distributed)

  FT of important signals: delta-distribution, Gaussian function, box function (To be distributed)

Signal sampling (To be distributed)

  Signal interpolation (To be distributed)
 

Introduction (Ch. 1.3-1.6)  
Spatial representation (Ch.  2.1-2.2)

Linear filtering

Sobel operator, 332-333

Separable filters, 109-111

Binomial filters 297-304

Counter example to linear filtering: median filtering 108-109

Multi-resolution (pp. 121-124 and pp. 134-138) Gaussian pyramid

Laplacian pyramid

Color (to be distributed)

Orientation (pp. 343-357)

Structure tensor and local orientation

Color display of orientation

Texture features (383-393, except section 12.3.2 ) Local Mean, Variance and Higher Moments

Local orientation as local linear symmetry feature

Laplacian (or Gaussian pyramid) pyramid as coarseness feature

Shape
 

Binary object representation: Chain code, Run-length code, Quad-tree code (pp. 502-507)

Simple shape parameters: Area, Perimeter, Circularity, Bounding box, (pp. 507-509)

Shape in binary images:  Fourier descriptors (pp. 512-516)

Shape in binary images: Erosion, Dilation, Thinning, Hit-Miss function, Boundary extraction. (pp. 489-502)

Shape in binary images:  Hough transform,  14.5.2 14.53   (pp. 462-464)
 
 

Classification (pp. 453-464, pp. 521-536)
  Feature selection

Dimensionality reduction

Classification techniques

16.3 Thresholding, Supervised Classification pp. 531-536

  14.1-14.3 Unsupervised Segmentation 453-461