计算机工程与应用 ›› 2015, Vol. 51 ›› Issue (11): 114-118.

• 数据库、数据挖掘、机器学习 • 上一篇    下一篇

支持线段查询索引结构CB树

徐红波1,2,姚念民2,韩启龙2,潘海为2   

  1. 1.哈尔滨商业大学 计算机与信息工程学院,哈尔滨 150028
    2.哈尔滨工程大学 计算机科学与技术学院,哈尔滨 150001
  • 出版日期:2015-06-01 发布日期:2015-06-12

Index structure CB-tree supporting line query

XU Hongbo1,2, YAO Nianmin2, HAN Qilong2, PAN Haiwei2   

  1. 1.College of Computer and Information Engineering, Harbin University of Commerce, Harbin 150028, China
    2.College of Computer Science and Technology, Harbin Engineering University, Harbin 150001, China
  • Online:2015-06-01 Published:2015-06-12

摘要: 在空间数据库中点、线段和区域是构成向量对象的三种基本实体。现有的索引结构能够将点或区域对象有效地组织成散列或分层目录,并且提供精确的检索方法。然而,这些索引结构索引线段时会出现以下问题。索引结构不能准确地表示线段的空间信息,这将阻碍对线段空间数据的高质量存储。位于层次目录中节点之间将产生大量死空间和重叠区域,随着时间的推移这将降低系统性能。提出一种采用数据压缩的索引结构CB树。与R树索引结构相比,CB树具有较优查询效率,占用较少存储空间。

关键词: 空间数据库, 线段, 索引结构, CB树

Abstract: Point, line and region are three basic entities which constitute vector-based objects in spatial databases. Existing index structures can effectively organize the objects such as points, regions into hash or hierarchical directory, and provide accurate retrieval methods. However the following questions arise when such methods are applied to line segments. Spatial information of line segments may not be precisely expressed which impedes high-quality conservation of line segments. In the hierarchical directory, a lot of dead space and overlapping regions will be generated between nodes, which degrade the performance over time. The paper presents the index structure CB-tree based on data compression. Compared with R-tree, CB-tree has optimum query efficiency, takes up less storage space.

Key words: spatial database, line segment, index structure, CB-tree