组最近邻查询是空间对象查询领域的一类重要查询,通过该查询可找到距离给定查询点集最近的空间对象。由于图像分辨率或解析度的限制等因素,空间对象的存在不确定性广泛存在于某些涉及图像处理的查询应用中。这些对象位置数据的存在不确定性会对组最近邻查询结果产生影响。本文给出面向存在不确定对象的概率闽值组最近邻查询定义,设计了高效的查询处理机制,通过剪枝优化等手段提高概率阈值组最近邻查询效率,并进一步提出了高效概率阈值组最近邻查询算法。采用多个真实数据集对概率阈值组最近邻算法进行了实验验证,结果表明所提算法具有良好的查询效率。
Group nearest neighbor query is an important type of spatial queries, the spatial object which is the nearest to the query point set can be found by this query. Due to low image resolution or the limitation of color definitions and so on, existentially uncertainty of spatial objects is inherent in the real applications involving image processing. The existentially uncertain locations will affect the re- suits of group nearest neighbor queries. This paper proposes the definition of probabilistic threshold group nearest neighbor (PTGNN) query over existentially uncertain data and designs an efficient query processing method. The efficiency of PTGNN query is improved by pruning methods and so on. An efficient algorithm is described in this paper. Extensive experiments based on real datasets have demonstrated the efficiency of the proposed algorithms.