Noblesse Workshop on Non-Linear Model Based Image Analysis

This book contains papers presented at the Noblesse Workshop on Non-linear model based image analysis held in Glasgow, 1-3 July 1998. Current models have mainly been developed for image coding purposes. They are rather simple and far away from being optimal and do not contribute to more complex tasks like those needed in image databases. This book meets the challenging tasks in multimedia applications by discussing new sophisticated model-based schemes for a high-level description of images and image sequences. Novel results are covered in the papers presented in this book, opening new potential fields of application like the support for building databases in multimedia applications, image archiving and image sequence coding, including such topics as:- 3D Image Models; Image/Video Restoration; Segmentation and Object Oriented Coding; Colour Image Processing; Database Retrieval; Image Models; Video Pre- and Post processing.

Contenu
3D Image Models.- Cross individual model-based alignment of volumetric images.- Model based segmentation and recognition of anatomical brain structures in 3D MR images.- Precise registration of CT and MR volumes based on a new creaseness measure.- Automatic object modeling for 3D virtual environment.- 3D binary fast morphological shape decomposition.- Stereo image segmentation using hybrid analysis technique.- A simplified representation of 3D human faces adapted from 2D images.- Active shape model-assisted least squares stereo matching for recovery of 3-D facial shape.- Parallel implementation of a 3D nonlinear restoration algorithm.- Keynote Lecture.- A perception-based error criterion and the design of stack filters for image restoration.- Image/Video Restoration.- RCRS filters for digital image sequences.- An automatic system for restoring image sequences.- Restoring images taken from scratched 2-inch tape.- Poster Session Segmentation and Object Oriented Coding.- Object recognition from a perimeter high curvature model using rank conditioned morphological operators.- Texture map coding for arbitrarily shaped 2-D and 3-D video objects.- Aligning arbitrary surfaces using pairwise geometric histograms.- Object tracking for retrival applications in MPEG-2.- Object feature based coding quality prediction for coding scheme selection.- Supervised semantic object segmentation and tracking via EM-based estimation of mixture density parameters.- Interactive object extraction from video sequences for multimedia applications based on multiple features.- Robust morphological scale-space trees.- Image segmentation and the use of genetic algorithms for optimising parameters and algorithm choice.- A class of parallel algorithms for nonlinear variational segmentation: A preprocessfor robust feature-based image coding.- Linear and non-linear modelling of scale-orientation signatures.- Parametric method for tracking and analysing lip movements.- Colour Image Processing (papers marked with a have colour illustrations contained in a separate Colour Annex).- Optimising lossless co-encoding of wavelet coefficients for entropy encoding.- Directional unsharp masking-based approach for color image enhancement.- Multidimensional texture characterization using moments.- Activity driven non-linear diffusion for colour image segmentation.- Old painting digital color restoration.- Multi-channel adaptive signal processing of colour cinematographic imagery.- Using colour for robust tracking and segmentation.- Keynote Lecture.- The trace transform and its applications.- Database Retrieval.- Image retrieval based on colour and nonlinear texture invariants.- Image matching via the inhomogeneous diffusion of color and texture features.- Shot detection tools in digital video.- Poster Session Image Models, Segmentation and Implementation.- Modeling of texture images by the structuring element of morphological filters.- Two dimensional volterra filter modelling of textured images using weighted constrained optimisation.- A retina-like image representation of primal sketch features extracted using a neural network approach.- Facial feature extraction and determination of pose.- A quantitative methodology for parameters setting in 3D people localization system for surveillance applications.- A segmentation methodology for real 3D images.- An optimal morphological non uniform sampling scheme for sequence image representation and compression.- 3D image analysis of facial shape changes using depth from stereo.- Dense disparity estimation via global and local matching.- Object segmentation and modeling in volumetric images.- Non linear image processing on field programmable gate arrays.- Video Pre and Post Processing (papers marked with a have colour illustrations contained in a separate Colour Annex).- Noise reduction techniques for video coding applications.- A robust hybrid bayesian-morphological approach to multidimensional image filtering.- Adaptively weighted vector-median processing of motion fields in block-DCT and object-based video coding.- Preprocessing for the MPEG-2 video coding.- Image interpolation by an integer factor using a non linear technique*.- On the structure of idempotent monotone Boolean functions.- Author index.