Acta Geodaetica et Cartographica Sinica ›› 2016, Vol. 45 ›› Issue (12): 1476-1484.doi: 10.11947/j.AGCS.2016.20160210

Previous Articles     Next Articles

A 3D Annotation Optimal Placement Algorithm for the Point Features in the Small Scale Geographic Scene

ZHOU Xinxin, WU Changbin, SUN Zaihong, DING Yuan, HE Tao   

  1. Key Laboratory of Virtual Geographic Environment, Ministry of Education, Nanjing Normal University, Nanjing 210046, China
  • Received:2016-05-03 Revised:2016-10-11 Online:2016-12-20 Published:2017-01-02
  • Supported by:
    The National Natural Science Foundation of China (No.41471318)

Abstract: The 3D annotations placement rules of point features in geographic scene are "obscured then not showing" and "obscured then directly showing" normally. The defects of those rules are annotation information lost or large numbers of occlusion, so their universalities are not strong and they are not suitable for the annotation placement of small-scaled geographic scene. This paper summarizes the contents, position and placement methods of 3D annotation and takes the aim at "not loss of annotation information and less annotation obscured as far as possible" of the research problem of the annotation placement of small-scaled geographic scene. The configuration rule of 3D annotation Identifies as "obscured then optimized to display". The designed algorithm based on the perspective transformation matrix, the inverse perspective transformation matrix and the grid algorithm takes the genetic algorithm (GA) whose fitness evaluation function uses the 3D Annotation quality evaluation function as the core to realize the feasible optimal solution of 3D annotations of point features in geographic scene. By the multi-views, multi-platforms contrast experiment, this algorithm is applicable for multi-views 3D annotation placement widely. The 3D annotation effect is better than mainstream GIS platforms (such as SuperMap desktop, ArcScene), which assumes that the algorithm's 3D annotation quality value is relatively increased 144%, 232%. The algorithm fits in with the target configuration.

Key words: geographic scene, 3D annotation, 3D annotation placement, genetic algorithm, 3D annotation quality evaluation function, GRID algorithm

CLC Number: