Syllabus
- image sampling and quantization, Shannon theorem, aliasing
- basic image operations, histogram, contrast stretching, noise removal, image sharpening
- linear filtering in the spatial and frequency domains, convolution, Fourier transform
- edge detection, corner detection
- image degradations and their modelling, inverse and Wiener filtering, restoration of motion-blurred and out-of-focus blurred images
- image segmentation
- image registration and matching
- features for description and recognition of 2-D shapes
- invariant features, Fourier descriptors, moment invariants, differential invariants
- statistical pattern recognition, supervised and nonsupervised classification, NN- classifier, linear classifier, Bayessian classifier
- clustering in a feature space, iterative and hierarchical methods
- dimensionality reduction of a feature space
Annotation
An introductory course on image processing and pattern recognition. Major attention is paid to image sampling and quantization, image preprocessing (noise removal, contrast stretching, sharpening, and de-blurring), edge detection, geometric transformations and warping, features for shape description and recognition, and to general pattern recognition techniques.
Numerous applications and experimental results are presented in addition to the theory.