460-4020/01 – Image Analysis (ANO)
| Gurantor department | Department of Computer Science | Credits | 5 |
| Subject guarantor | doc. Dr. Ing. Eduard Sojka | Subject version guarantor | doc. Dr. Ing. Eduard Sojka |
| Study level | undergraduate or graduate | | |
| | Study language | Czech |
| Year of introduction | 2010/2011 | Year of cancellation | 2014/2015 |
| Intended for the faculties | FEI | Intended for study types | Follow-up Master |
Subject aims expressed by acquired skills and competences
The course acquaints the students with the fundamental methods of image anlaysis.
Teaching methods
Lectures
Seminars
Summary
The following topics are discussed: Image segmentation. Detecting edges, regions, corners. Measuring objects. Pattern recognition based on classification. Using neural networks for pattern recognition. Analysis of images of 3D scenes. Processing images varying in time. Tracking objects.
Compulsory literature:
Recommended literature:
Additional study materials
Way of continuous check of knowledge in the course of semester
Conditions for credit:
The project and the tasks assigned during the exercises must be worked out.
E-learning
Other requirements
Additional requirements are placed on the student
Prerequisities
Subject has no prerequisities.
Co-requisities
Subject has no co-requisities.
Subject syllabus:
Lectures:
Detecting edges in images. Gradient, and zero-crossing methods. Parametric edge models.
Canny edge detector.
Thresholding. Optimal threshold selection.
Image segmentaion based on region growing and splitting.
Edge linking. Heuristic edge following. Method of clustering in the space of parameters.
Representing boundaries and areas: Encoding boundaries using line segments and curves. Line and curve fitting. Representing areas.
Detecting feature points (corners).
Measuring objects. Selection and computation of features for pattern recognition. Evaluating the efficiency and optimisation of the set of selected features.
Pattern recognition based on classification. Discriminant functions and etalons.
Probablistic approach to determining the discriminant functions.
Using neural networks for pattern recognition.
Reconstructing a scene from its two or more images.
Absolute and relative camera calibration and reconstruction.
Analysis of time-varying images. Tracking objects in image sequences.
Projects:
As a project, the students will carry out a simple recogniser of two-dimensional geometrical shapes.
Computer labs:
During the exercises, the students work out a series of practical tasks (detecting edges, areas, corners, computing features, etalon-based classification). The tasks are prepared in the form of templates (pre-prepared programs) into which the sudents fill their own source code. In this way, they can focus on substantial and interesting issues. Furthermore, examples of reconstructing 3D scenes are presented.
Conditions for subject completion
Conditions for completion are defined only for particular subject version and form of study
Occurrence in study plans
Occurrence in special blocks
Assessment of instruction