电讯技术2018，Vol.58Issue(2)：174-177,4.DOI:10.3969/j.issn.1001-893x.2018.02.010
基于无监督学习的目标轨迹预测
Target Trajectory Prediction Based on Unsupervised Learning
成磊峰1
作者信息
- 1. 中国西南电子技术研究所,成都610036
- 折叠
摘要
Abstract
Aiming at the problem of target trajectory prediction in the field of intelligence and reconnais-sance surveillance,a prediction method based on unsupervised learning is proposed.Firstly,a target histori-cal activity pattern is analyzed according to historical information.Secondly,the hidden Markov model is constructed and the motion direction of the prediction target in the grid is automatically realized by unsu-pervised learning.Finally,The target trajectory in the short term is calculated according to the learning di-rection of motion and the speed of target motion.Numerical simulations show that the proposed method can effectively predict the trajectory of the target in the short time (usually 5 minutes).关键词
情报侦察监视/目标轨迹预测/目标历史信息/活动规律分析/无监督学习/栅格化Key words
intelligence and reconnaissance surveillance/target trajectory prediction/target historical information/activity pattern analysis/unsupervised learning/rasterize分类
信息技术与安全科学引用本文复制引用
成磊峰..基于无监督学习的目标轨迹预测[J].电讯技术,2018,58(2):174-177,4.
