A Mathematical Morphology Based Scale Space Method for the Mining of Linear Features in Geographic Data

Authors:
Min Wang;Yee Leung;Chenhu Zhou;Tao Pei;Jiancheng Luo
Affiliations:
College of Geography Science, Nanjing Normal University, Nanjing, China;Department of Geography and Resource Management, Center for Environmental Policy and Resource Management, and Joint Laboratory for Geoinformation Science, The Chinese University of Hong Kong, Shat ...;State Key Laboratory of Resources and Environment Information System, Chinese Academy of Sciences, Beijing, China;State Key Laboratory of Resources and Environment Information System, Chinese Academy of Sciences, Beijing, China;State Key Laboratory of Resources and Environment Information System, Chinese Academy of Sciences, Beijing, China
Venue:
Data Mining and Knowledge Discovery
Year:
2006

Citing 10
Cited 1

Pattern Spectrum and Multiscale Shape Representation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Cluster Analysis by Binary Morphology

IEEE Transactions on Pattern Analysis and Machine Intelligence
Clustering data by melting

Neural Computation
Variants for the Hough transform for line detection

Computational Geometry: Theory and Applications
Scale-Space Using Mathematical Morphology

IEEE Transactions on Pattern Analysis and Machine Intelligence
Clustering by Scale-Space Filtering

IEEE Transactions on Pattern Analysis and Machine Intelligence
Density-Based Clustering in Spatial Databases: The Algorithm GDBSCAN and Its Applications

Data Mining and Knowledge Discovery
A Fast Parallel Clustering Algorithm for Large Spatial Databases

Data Mining and Knowledge Discovery
Image Analysis and Mathematical Morphology

Image Analysis and Mathematical Morphology
Scale space classification using area morphology

IEEE Transactions on Image Processing

Delineation of support domain of feature in the presence of noise

Computers & Geosciences

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a spatial data mining method MCAMMO and its extension L_MCAMMO designed for discovering linear and near linear features in spatial databases. L_MCAMMO can be divided into two basic steps: first, the most suitable re-segmenting scale is found by MCAMMO, which is a scale space method with mathematical morphology operators; second, the segmented result at this scale is re-segmented to obtain the final linear belts. These steps are essentially a multi-scale binary image segmentation process, and can also be treated as hierarchical clustering if we view the points under each connected component as one cluster. The final number of clusters is the one which survives (relatively, not absolutely) the longest scale range, and the clustering which first realizes this number of clusters is the most suitable segmentation. The advantages of MCAMMO in general and L_MCAMMO in particular, are: no need to pre-specify the number of clusters, a small number of simple inputs, capable of extracting clusters with arbitrary shapes, and robust to noise. The effectiveness of the proposed method is substantiated by the real-life experiments in the mining of seismic belts in China.