Adjoint coefficient calculation method applied to trajectory adjoint possibility analysis

Abstract

The invention discloses an adjoint coefficient calculation method applied to trajectory adjoint possibility analysis. The adjoint coefficient calculation method comprises the steps of obtaining two trajectories A and B and sorting the trajectories A and B according to time; traversing a track point P of the track A, finding out a track point Q closest to the track point P in time from the track B,and calculating and recording a time difference of a track point pair (P, Q); finding out a track point pair (P ', Q') which has the smallest time difference and meets a given condition from the track point pairs (P, Q); dividing the track A into a track A1 and a track A2 by a track point P ', and dividing the track B into a track B1 and a track B2 by a track point Q'; repeating the above steps until all matched track point pairs (P ', Q') are found, and storing the track point pairs in an optimal matching result set; calculating and counting a matching state; calculating adjoint coefficients. According to the method, the calculation complexity of the adjoint coefficient is effectively reduced, the calculation efficiency is improved, the influence of the deviation value is small, and themethod can be applied to actual trajectory data to analyze effective adjoint information.

Description

technical field [0001] The invention relates to security and information, in particular to a method for calculating an adjoint coefficient applied to the analysis of the adjoint possibility of a trajectory. Background technique [0002] At present, a variety of monitoring equipment in the security system will collect a large amount of behavior trajectory data, the behavior trajectory data includes data such as face, MAC, IMSI and IMEI, and different types of data are usually collected by different types of collection equipment. Therefore, different types The track data of the track usually has no direct relationship, but a lot of useful information can be analyzed from the above track data through data mining. One of the important information is the accompanying state of the track. Through the accompanying state, it is possible to find the accompanying Personnel, or identifying multiple behavioral characteristics belonging to the same target person, information such as this ...

AI Technical Summary

Problems solved by technology

[0003] For the accompanying analysis of the trajectory, the trajectory of the accompanying person is not necessarily completely consistent with the target person, but the two are at an appropriate distance within a certain time difference, but the time difference and distance are not fixed and often change, and sometimes there will be deviations. The accompanying state may be discontinuous (that is, the trajectory of the accompanying person and the trajectory of the target person are separated after a period of time, and then return to the accompanying state after a period of time), these problems need to be considered when doing trajectory accompanying analysis; in addition, The adjoint analysis of the trajectory needs to calculate a large amount of data to obtain the adjoint coefficient. There are many trajectory data analysis algorithms for calculating the adjoint coefficient in the market, but most of the trajectory data analysis algorithms are too complicated and time-consuming to be effectively applied to large-scale trajectories. in the data

Images