Author:
Sun Mengzhu,Wang Jiasheng
Abstract
Abstract
A clustering method based on Minimum Bounding Rectangle and buffer similarity was proposed to improve the trajectories clustering and reduce computation time. Taking the Volunteer Observation Ship data as an example, firstly, we calculated similarity by dividing the area of the overlapped parts of two MBRs by the area of the larger MBR. According to the similarity matrix, the original trajectories were divided into several clusters by coarse clustering using Density-Based Spatial Clustering of Applications with Noise (DBSCAN). After that, the similarity of each cluster was calculated by buffer analysis. Then DBSCAN was used again for each cluster, and the Silhouette Coefficient was used to evaluate the resulting cluster quality. The experimental results showed that the presented methods improve the accuracy of MBR and reduce time cost of buffer similarity.
Reference23 articles.
1. Safety assessment of shipping routes in the South China Sea based on the fuzzy analytic hierarchy process;Jiasheng;Safety. Sci.,2014
2. A new method for discovering behavior patterns among animal movements;Yuwei;Int. J. Geogr. Inf. Sci.,2015
3. The use of trajectory cluster analysis to interpret trace gas measurements at Mace Head, Ireland;Cape;Atmos. Environ.,2000
4. Review of the Research Progresses in Trajectory Clustering Methods;Xi;Progress in Geography,2011
5. Using derivatives in a longest common subsequence dissimilarity measure for time series classification;Górecki;Pattern Recogn. Let.,2014
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献