摘要
热释电红外(Passive Infrared,PIR)传感器人体行为研究目前主要集中于特征建模与识别,对轨迹重构的研究相对较少且易忽略传感器布设环境的影响,导致重构结果不确定性较大。为提高重构轨迹的完备性,该文将传感器与环境的连通性作为PIR传感器网络轨迹重构中的约束条件,从空间相邻、时间连续、传感器外连通三方面对轨迹重构过程进行优化。以PIR传感器系统模拟场景中对象随机移动为例,对传感器的响应数据进行轨迹重构,得到对象的连续行为轨迹,结果与场景模拟轨迹基本一致,验证了该文方法的有效性。
Passive infrared(PIR)sensors have been widely used in the analysis of human behavior,such as motion analysis,indoor tracking,and intelligent security.Currently,research in this field primarily focuses on feature modeling and recognition,while research on trajectory reconstruction is relatively limited.Existing trajectory reconstruction methods overlook the impact of sensor distribution environments,resulting in high uncertainty in the reconstruction results.In order to improve the completeness of the reconstructed trajectories,this paper considers the connectivity between sensors and the environment as a constraint in PIR sensor network trajectory reconstruction.The optimization of the trajectory reconstruction process is carried out from three aspects:spatial adjacency,temporal continuity and external connectivity of the sensors.Taking the example of random movement of objects in a PIR sensor system simulation scenario,this paper reconstructs the continuous behavioral trajectories of the objects using response data from the sensors.Through a comparison between the reconstructed trajectories and the actual trajectories in the scenario,it is found that the two sets of trajectories are generally consistent,which confirms the effectiveness of the proposed method.
作者
潘炳煌
滕玉浩
钱凌欣
罗文
俞肇元
PAN Binghuang;TENG Yuhao;QIAN Lingxin;LUO Wen;YU Zhaoyuan(Key Laboratory of Virtual Geographic Environment,Nanjing Normal University,Ministry of Education,Nanjing 210023;State Key Laboratory Cultivation Base of Geographical Environment Evolution(Jiangsu Province),Nanjing 210023;Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application,Nanjing 210023,China)
出处
《地理与地理信息科学》
CSCD
北大核心
2023年第5期1-7,共7页
Geography and Geo-Information Science
基金
国家自然科学基金项目(41971404、42130103)
江苏省研究生科研与实践创新计划项目(SJCX22_0533)。
作者简介
潘炳煌(1998-),男,硕士研究生,主要从事路网轨迹数据索引相关研究;通信作者:俞肇元,E-mail:yuzhaoyuan@njnu.edu.cn。
