Course detail

Analysis of Signals and Images

FEKT-LASOAcad. year: 2020/2021

Time-frequency signal analysis. Continuous and discrete image representation, 2D transforms, stochastic image. Enhancement and edition of images - contrast transforms, sharpening, noise and interference suppression, geometric operations. Introduction to restoration of distorted images. Methods of image reconstruction from parallel and fan tomographic projections. Non-linear analysis and filtering of signals and images, neuronal classifiers. Edge, border and area detection, image segmentation. Analysis and visualisation of 2D and 3D image data. Technical, medical and ecological applications.

Language of instruction

Czech

Number of ECTS credits

5

Mode of study

Not applicable.

Learning outcomes of the course unit

The graduate is capable of:
- being oriented in theoretical principles of signal and image analysis methods, and also in practical aspects of their implementation,
- designing suitable approaches and also provide consultations in this respect,
- aplying the respective programmes including commercial software and also of programming independently designed related algorithms,
- being a valid member of intedisciplinary teams in the area of signal and namely image analysis.

Prerequisites

The subject knowledge on the Bachelor´s degree level is requested, particularly mathematics and digital signal processing

Co-requisites

Not applicable.

Planned learning activities and teaching methods

Teaching methods depend on the type of course unit as specified in the article 7 of BUT Rules for Studies and Examinations. Techning methods include lectures and computer laboratories, and namely selfstudy. Course is taking advantage of e-learning (Moodle) system. Students have to write projects/assignments during the course.

Assesment methods and criteria linked to learning outcomes

Requirements for completion of a course are elaborated by the lecturer responsible for the course every year;
basically:
- obtaining at least 15 points (out of 30 as course-unit credit based on active presence in demonstration exercises),
- successful passing of final written exam (up to 70 points)

Course curriculum

1. Time-frequency signal analysis, wavelet transforms.
2. Continuous image representation, 2D transforms, stochastic image.
3. Discrete and digital image representation, 2D discrete transforms, discrete operators.
4. Enhancement and editing of images – contrast and colour scale transforms.
5. Mask operators, sharpening, noise suppression, geometric operations.
6. Introduction to restoration of distorted images
7. Local parameters, texture analysis and parametric image.
8. Image segmentation based on homogeneity, region oriented segmentation.
9. Image segmentation based on edge representation, Hough transform.
10. Image segmentation by the watershed method. Segmentation by flexible contours and level sets.
11. Generalised morphological transforms.
12. Reconstruction methods of images from parallel and fan projections, in original and spectral domain.
13 Nonlinear analysis and filtering of images, neuronal classifiers.

Work placements

Not applicable.

Aims

The goal of the course is to provide the students with knowledge of time-frequency signal analysis and particularly of digital signal processing and analysis.

Specification of controlled education, way of implementation and compensation for absences

Delimitation of controlled teaching and its procedures are specified by a regulation issued by the lecturer responsible for the course and updated for every year (see Rozvrhové jednotky).
Basically:
- obligatory computer-lab tutorial
- voluntary lecture

Recommended optional programme components

Not applicable.

Prerequisites and corequisites

Not applicable.

Basic literature

J.Jan: Číslicová filtrace, analýza a restaurace signálů. VUTIUM 2002
J.Jan: Medical Image Processing, Reconstruction and Restoration. CRC 2006

Recommended reading

Not applicable.

eLearning

Classification of course in study plans

  • Programme EEKR-ML1 Master's

    branch ML1-BEI , 1. year of study, winter semester, compulsory

Type of course unit

 

Lecture

39 hours, optionally

Teacher / Lecturer

Syllabus

Time-frequency signal analysis, wavelet transforms
Continuous image representation, linear 2D systems, 2D spectra
Discrete image representation, discrete 2D linear operators, separable and convolutory local operators
Discrete 2D transforms: DFT, cosine, sine, Hadamard, Haar transforms
Image enhancement, edition and geometric operations
Image noise and interference suppression
Elements of formalised image restoration, pseudoinverse filtering
Tomographic image reconstructions from projections - principles of algebraic methods, of methods via frequency domain and by filtered back-projection
Non-linear analysis of signals and images - homomorphic and median filtering
Elements of signal/image analysis and filtering by neural networks
Image segmentation, edge, boarder and area detection
Movement- and depth analysis. Visualisation of 3D and 4D image data.
Applications of image analysis in technology, medicine and ecology.

Exercise in computer lab

13 hours, compulsory

Teacher / Lecturer

Syllabus

Becoming acquainted with MATLAB - Image Processing Toolbox environment
Wavelet analysis of complicated signals
Experimental acquisition of image data. Basic operations with image data in original domain
2D discrete systems, verification of characteristics
Generating of discrete stochastic fields, correlation analysis
2D DFT, image spectra
Contrast end colour enhancement, histogram equalisation
Image sharpening and interference suppression
Distortion (or blurr) identification, design and verification of modified inverse filtering
Experimental Radon transform and aproximative reconstruction from projections based on spectral slice theorem
Aproximative reconstruction from projections by filtered back projection
Basic methods of image segmentation, texture analysis
Manipulation with image data in common compressed formats

eLearning