Object trajectory similarity obtaining method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a method for obtaining object track similarity. The method comprises the following steps: obtaining a spatial-temporal characteristic matrix corresponding to each target object; performing dimension reduction operation on the spatial-temporal characteristic matrix to obtain a signature matrix corresponding to the spatial-temporal characteristic matrix; performing similarity calculation on the signature matrixes corresponding to any two spatial-temporal feature matrixes to obtain the signature matrix similarity between any two target objects; and selecting two target objects of which the signature matrix similarity is greater than a preset similarity threshold, and determining the trajectory similarity between the two selected target objects based on the spatial-temporal feature matrixes corresponding to the two selected target objects. According to the method, dimension reduction operation is carried out on the target spatial-temporal characteristic matrix corresponding to the target object, so that the dimension of data is greatly reduced, the data volume is greatly reduced, the track similarity obtaining efficiency of the target object is improved, the tracking response speed of the tracked object is high, and the real-time performance is good.

Description

technical field [0001] The invention relates to the field of data management, in particular to a method, device, electronic equipment and storage medium for obtaining object trajectory similarity. Background technique [0002] From the historical trajectory, extract the eigenvalues ​​of time and space to form a 24-dimensional time vector and an N-dimensional space vector. After space-time fusion, a 24*N-dimensional matrix is ​​formed, which is the space-time characteristic matrix. Among them, the N-dimensional space vector It needs to be determined according to the size of the actual space. [0003] At present, in the public security scene, the trajectory data of the tracked suspect (tracked object) is usually the above-mentioned spatio-temporal feature matrix, the data volume of the trajectory data is extremely large, and the large-scale and high-dimensional trajectory data challenges the performance of calculations. Real-world scenarios All of them are billion-level data,...

AI Technical Summary

Problems solved by technology

[0004] In related technologies, in order to save costs, computer equipment with weak performance is still selected for data calculation, but this also leads to low data calculation efficiency of the tracked object, making the response speed of the tracked object slow and poor real-time

Images