A Framework for Bus Trajectory Extraction and Missing Data Recovery for Data Sampled from the Internet

This paper presents a novel framework for trajectories’ extraction and missing data recovery for bus traveling data sampled from the Internet. The trajectory extraction procedure is composed of three main parts: trajectory clustering, trajectory cleaning and trajectory connecting. In the clustering...

Full description

Saved in:
Bibliographic Details
Published in:Sensors (Basel, Switzerland) Vol. 17; no. 2; p. 342
Main Authors: Tong, Changfei, Chen, Huiling, Xuan, Qi, Yang, Xuhua
Format: Journal Article
Language:English
Published: Switzerland MDPI AG 10.02.2017
MDPI
Subjects:
Algorithms
anomaly detection
Clustering
Data recovery
Extraction
Fuzzy
Internet
Missing data
missing data recovery
Trajectories
trajectory processing
anomaly detection
clustering
missing data recovery
trajectory processing
ISSN:1424-8220, 1424-8220
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper presents a novel framework for trajectories’ extraction and missing data recovery for bus traveling data sampled from the Internet. The trajectory extraction procedure is composed of three main parts: trajectory clustering, trajectory cleaning and trajectory connecting. In the clustering procedure, we focus on feature construction and parameter selection for the fuzzy C-means clustering method. Following the clustering procedure, the trajectory cleaning algorithm is implemented based on a new introduced fuzzy connecting matrix, which evaluates the possibility of data belonging to the same trajectory and helps detect the anomalies in a ranked context-related order. Finally, the trajectory connecting algorithm is proposed to solve the issue that occurs in some cases when a route trajectory is incorrectly partitioned into several clusters. In the missing data recovery procedure, we developed the contextual linear interpolation for the cases of missing data occurring inside the trajectory and the median value interpolation for the cases of missing data outside the trajectory. Extensive experiments are conducted to demonstrate that the proposed framework offers a powerful ability to extract and recovery bus trajectories sampled from the Internet.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:1424-8220
1424-8220
DOI:10.3390/s17020342