Traditional nearest neighbor queries and its variants, such as
Group Nearest Neighbor Query
(GNN), have been widely studied by many researchers. Recently obstacles are involved in spatial queries. The existence of obstacles may affect the query results due to the visibility of query point. In this paper, we propose a new type of query,
Group Visible Nearest Neighbor Query
(GVNN), which considers both visibility and distance as constraints.
Multiple Traversing Obstacles
(MTO) Algorithm and
Traversing Obstacles Once
(TOO) Algorithm are proposed to efficiently solve GVNN problem. TOO resolves GVNN by defining the invisible region of MBR of query set to prune both data set and obstacle set, and traverses obstacle R*-tree only once. The experiments with different settings show that TOO is more efficient and scalable than MTO.