Vehicle trajectory clustering method based on bag-of-word model and metric learning

Abstract

The invention discloses a vehicle trajectory clustering method based on a bag-of-word model and metric learning. The method comprises the following steps: extracting vehicle trajectory point motion parameters; dividing the vehicle track into a plurality of optimal homogeneous vehicle track segments based on a minimum description length principle, and obtaining statistical features of each segment to obtain a vehicle track segment feature set; coding each optimal homogeneous vehicle track segment into a fixed-length vector based on a bag-of-word model and a non-convex low-rank constraint metric learning method so as to obtain a feature descriptor of the optimal homogeneous vehicle track segment; and gathering the feature descriptors through a K-means algorithm and the non-convex low-rank constraint metric learning method to realize vehicle trajectory clustering. According to the method, speed, acceleration and steering angle information is introduced, feature coding is carried out on vehicle tracks, and then the vehicle tracks in different time periods and areas are gathered; in addition, a non-convex low-rank constraint-based metric learning method is provided to improve the similarity measurement performance.

Description

technical field [0001] The invention relates to a vehicle trajectory clustering method, in particular to a vehicle trajectory clustering method based on a bag of words model and metric learning. Background technique [0002] With the rapid development of global positioning system (GPS, Global Positioning System), wireless sensor network (WSN, Wireless Sensor Network) and other related positioning technologies, it is easier to track vehicles, and their motion trajectories can be acquired and processed to mine the underlying information. related physical properties, etc. As an important part of the spatio-temporal trajectory mining task, the clustering of vehicle trajectories gathers similar trajectories to form trajectory clusters to characterize the potential behavior characteristics of vehicles and predict the location and detect abnormalities accordingly. Therefore, it is of great significance in practical applications. [0003] Clustering of vehicle trajectories is usual...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the poor clustering effect of the traditional vehicle trajectory clustering method based on distance measurement, the purpose of the present invention is to provide a vehicle trajectory clustering method based on the bag-of-words model and metric learning, which is based on the bag-of-words model theory and according to the correlation of vehicle trajectory The physical characteristics encode the characteristics of the trajectory to obtain a fixed-length vehicle trajectory feature descriptor, and then use the improved K-means algorithm to gather the feature descriptors of all vehicle trajectories, and then form different trajectory clusters to represent the behavior characteristics of the vehicle, so as to detect Abnormal Vehicle Behavior and Predicted Vehicle Location

Images