DocumentCode
2462059
Title
Minimum description length based 2D shape description
Author
Li, Mengxiang
Author_Institution
Comput. Vision & Active Perception Lab., R. Inst. of Technol., Stockholm, Sweden
fYear
1993
fDate
11-14 May 1993
Firstpage
512
Lastpage
517
Abstract
The problem of 2-D shape description, particularly with contour partitioning, grouping, and classification in terms of straight and curved, based on the minimum description length (MDL) criterion and shape-fitting techniques, is discussed. The MDL criterion is used to detect outliers in connection with shape fitting. Using the MDL criterion, it is possible to derive for a given data set and a class of models a description which best explains the data. A new algorithm for fitting 2-D points to an ellipse is presented
Keywords
computer vision; image classification; 2-D points; 2-D shape description; classification; contour partitioning; ellipse; grouping; minimum description length; outliers; Computer vision; Curve fitting; Information theory; Laboratories; Maximum likelihood estimation; Parametric statistics; Probability; Shape; Stochastic processes; Vocabulary;
fLanguage
English
Publisher
ieee
Conference_Titel
Computer Vision, 1993. Proceedings., Fourth International Conference on
Conference_Location
Berlin
Print_ISBN
0-8186-3870-2
Type
conf
DOI
10.1109/ICCV.1993.378170
Filename
378170
Link To Document