CONSTANT VELOCITY SEARCH ALGORITHM FOR TRAJECTORY QUERY BASED ON REGION OF INTEREST

摘要

A constant velocity search algorithm for trajectory query based on a region of interest, comprising the following steps: step 1: initially setting the lower limit of global spatial density correlation LB＝0 and the upper limit of the global spatial density correlation UB=+∞; step 2: extending a point within the smallest distance each time from each center of query regions by using a Dijkstra extension algorithm of the same rate; step 3: calculating the upper limit and the lower limit of spatial density to update LB and UB; step 4: determining whether LBUB or a search radius greater than ε is true, ε being a preset threshold, if yes, ending the search, and entering the next step, i.e., step 5; if not, returning to step 3; step 5: sorting trajectories according to the upper limit value of the spatial density; step 6: performing further refinement according to the trajectory sorting, and returning the trajectory having maximum spatial density correlation. Different from traditional trajectory search (TSL) query, the algorithm is based on a region of interest, and resolves the problem invalidity of existing TSL solutions valid for TSR query.