Minimum Description Length Based 2 D Shape DescriptionMengxiang
نویسنده
چکیده
In this paper, we present an approach for 2{D shape description based on the Minimum-Description-Length (MDL) criterion. Using this criterion, we can derive, for a given data set and a class of models, a description which best explains the data. The problems of contour partitioning, colinear and cocurvilinear segment grouping, and classiication of 2{D contours in terms of straight and curved, are formulated in a particular description language and the MDL criterion is applied, the shortest description is accepted as the \explanation" of the data. The algorithm has been implemented and tested on a variety of real images, giving reasonable results. The work presented in the paper was performed under the ESPRIT-BRA 3038 Vision as Process (VAP) project. The support from the Swedish National Board for Industrial and Technical Development, NUTEK, is gratefully acknowledged. Discussion with and comments from Jan-Olof Eklundh, GG oran Olofsson, and Lars Olsson are very much appreciated.
منابع مشابه
Add Cartesian Differential Invariants to Minimum Description Length Shape Models
The minimum description length approach can automatic solve the point correspondence problem and give the better statistical shape models than those built by hand or equally spaced way. The current mdl approaches build the models only based on the segmented shapes without considering the local image structure and may place the markers at wrong places. This paper adds Cartesian differential inva...
متن کاملMDL Shape Models: Implementation Issues
Minimum Description Length (MDL) shape modelling has recently established itself as a gold standard in optimal shape modelling. By end of 2004, available MDL packages are written in Matlab and/or incorrectly. The purpose of this technical report is to list the most critical problems that occurred during a C++ implementation of MDL–based 2D shape modelling.
متن کاملDeformable Shape Detection and Description via Model-Based Region
A method for deformable shape detection and recognition is described. Deformable shape templates are used to partition the image into a globally consistent interpretation, determined in part by the minimum description length principle. Statistical shape models enforce the prior probabilities on global, parametric deformations for each object class. Once trained, the system autonomously segments...
متن کاملBuilding Reconstruction Based on Mdl Principle from 3-d Feature Points
For 3-D building reconstructions of urban areas, we present a fully automatic shape recovery method that uses 3-D points acquired from aerial image sequences. This paper focuses on shape recovery of flat rooftops that are parallel to the ground. We recover each rooftop from a set of 3-D points located at nearly the same height. Such 3-D point sets are made by merging point sets under the MDL (M...
متن کاملRegion Segmentation via Deformable Model-Guided Split and Merge
An improved method for deformable shape-based image segmentation is described. Image regions are merged together and/or split apart, based on their agreement with an a priori distribution on the global deformation parameters for a shape template. Perceptually-motivated criteria are used to determine where/how to split regions, based on the local shape properties of the region group’s bounding c...
متن کامل