Ground maneuvering target motion trail semantic system construction and cognitive optimization method

Abstract

The invention relates to the field of pattern recognition and geographic information, and discloses a ground maneuvering target motion trail semantic system construction and cognitive optimization method. Basic motion semantics are defined according to track characteristics corresponding to different motion modes, and a ground maneuvering target motion track semantic system is constructed. And a ground maneuvering target motion trail semantic body knowledge strategy is optimized, and the algorithm operation time is shortened. And an identification matching method of a target track and a motion mode is provided, and intelligent cognition of target motion behavior semantics is realized.

Description

technical field [0001] The present invention relates to the fields of pattern recognition and geographic information, and in particular to a method for constructing a semantic system of a movement track of a ground maneuvering target and a cognitive optimization method. Background technique [0002] The trajectories of ground maneuvering targets have certain rules. Trajectory analysis can help to mine the semantics of their motion behaviors and understand the target’s intentions, explore the rules of target behavior, and then assist in decision-making. In recent years, the research directions of ground maneuvering target trajectory data analysis mainly include position prediction, interest mining, path planning, etc., and trajectory semantic mining is an emerging hotspot in trajectory analysis. [0003] However, the existing trajectory semantic mining schemes mainly face the following problems: the current trajectory semantic mining research is more focused on exploring and ...

AI Technical Summary

Problems solved by technology

[0003] However, the existing trajectory semantic mining schemes mainly face the following problems: the current trajectory semantic mining research is more focused on exploring and verifying better algorithm models on open source datasets, summarizing relevant motion laws and performing pattern recognition for specific datasets, and has not Considering multiple sources of trajectory data, it is difficult to adapt to the temporal and spatial complexity of the environment

Existing research on object trajectory mining and semantic modeling of motion behavior is relatively independent, and has little contribution to the construction of semantic system. The span of the trajectory studied is relatively limited in time and space, and most of them can only be applied to specific scenarios, with a narrow coverage , it is difficult to cope with the high complexity and mode diversity shown by a large number of target trajectories, and the related motion feature extraction and mining work still needs to be improved

The existing trajectory semantic mining schemes pay more attention to the accuracy of pattern recognition and the accuracy of semantic extraction results, and the optimization of computing time depends more on the improvement of hardware level. There are few studies on optimizing cognitive strategies to reduce computing time in principle. , there is still much room for improvement

Images