Voronoi classfied and clustered constellation data structure for three-dimensional urban buildings
In the past few years, the growth of urban area has been increasing and has resulted immense number of urban datasets. This situation contributes to the difficulties in handling and managing issues related to urban area. Huge and massive datasets can degrade the performance of data retrieval and inf...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2017
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/78844/1/NorSuhaibahAzriPFGHT2017.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In the past few years, the growth of urban area has been increasing and has resulted immense number of urban datasets. This situation contributes to the difficulties in handling and managing issues related to urban area. Huge and massive datasets can degrade the performance of data retrieval and information analysis. In addition, urban environments are very difficult to manage because they involved with various types of data, such as multiple types of zoning themes in urban mixeduse development. Thus, a special technique for efficient data handling and management is necessary. In this study, a new three-dimensional (3D) spatial access method, the Voronoi Classified and Clustered Data Constellation (VOR-CCDC) is introduced. The VOR-CCDC data structure operates on the basis of two filters, classification and clustering. To boost up the performance of data retrieval, VORCCDC offers a minimal percentage of overlap among nodes and a minimal coverage area in order to avoid repetitive data entry and multi-path queries. Besides that, VOR-CCDC data structure is supplemented with an extra element of nearest neighbour information. Encoded neighbouring information in the Voronoi diagram allows VOR-CCDC to optimally explore the data. There are three types of nearest neighbour queries that are presented in this study to verify the VOR-CCDC’s ability in finding the nearest neighbour information. The queries are Single Search Nearest Neighbour query, k Nearest Neighbour (kNN) query and Reverse k Nearest Neighbour (RkNN) query. Each query is tested with two types of 3D datasets; single layer and multi-layer. The test demonstrated that VOR-CCDC performs the least amount of input/output than their best competitor, the 3D R-Tree. Besides that, VOR-CCDC is also tested for performance evaluation. The results indicate that VOR-CCDC outperforms its competitor by responding 60 to 80 percent faster to the query operation. In the future, VOR-CCDC structure is expected to be expanded for temporal and dynamic objects. Besides that, VOR-CCDC structure can also be used in other applications such as brain cell database for analysing the spatial arrangement of neurons or analysing the protein chain reaction in bioinformatics applications. |
|---|