Intelligent transportation systems data compression using wavelet decomposition technique.
-
2009-12-01
-
Details:
-
Creators:
-
Corporate Creators:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Corporate Publisher:
-
Abstract:Intelligent Transportation Systems (ITS) generates massive amounts of traffic data, which posts
challenges for data storage, transmission and retrieval. Data compression and reconstruction technique plays an
important role in ITS data procession. Traditional compression methods have been utilized in Transportation
Management Centers (TMCs), but the data redundancy and compression efficiency problems remain. In this
report, the wavelet incorporated ITS data compression method is initiated. The proposed method not only
makes use of the conventional compression techniques but, in addition, incorporates the one-dimensional
discrete wavelet compression approach. Since the desired wavelet compression is a lossy algorithm, the
balancing between the compression ratio and the signal distortion is exceedingly important. During the
compression process, the determination of the threshold is the key issue that affects both the compression ratio
and the signal distortion. An algorithm is proposed that can properly select the threshold by balancing the two
contradicted aspects. Three performance indexes are constructed and the relationships between the three
indices and the threshold are identified in the algorithm. A MATLAB program with the name Wavelet
Compression for ITS Data (WCID) has been developed to facilitate the compression tests. A case study on
TransGuide ITS data was put into play and a final compression ratio of less than one percent on the trade-off
threshold value shows that the proposed approach is practical. Finally, the threshold selection algorithm can be
further tuned up utilizing Autoregressive model so that the quality of reconstructed data can be improved with
a minor overhead of saving only a few parameters.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: