期刊文献+

物联网环境下异构传感数据融合目标识别算法 被引量:3

Target Recognition Algorithm for Heterogeneous Sensor Data Fusion in the Internet of Things Environment
下载PDF
导出
摘要 在对运动目标进行识别过程中,不同视角使运动目标姿态多变,传统算法的单一特征描述目标能力有限,导致识别精度下降,因此设计一种物联网环境下异构传感数据融合目标识别算法。首先建立异构数融合模型,通过关联处理剔除冗余数据,利用转换坐标系的方法完成多传感器的空间对准;利用直方图特征法结合二维傅里叶变换,提高相位谱图像对比度,取得合理的灰度数量级,完成目标特征提取;最后改进特征之间的冲突系数,将冲突系数超过阈值的特征进行处理,降低高冲突特征对整个识别流程的影响,至此完成目标识别算法的设计。通过仿真实验结果表明,设计的算法识别正确率与标准差的结果均优于传统方法,验证了设计算法的可靠性。 In the process of recognizing moving targets,different perspectives make the moving targets postures changeable.The single feature of traditional algorithms has limited ability to describe the target,resulting in a decrease in.identification accuracy.Therefore,a target recognition algorithm for heterogeneous sensor data fusion in the Internet of things environment was designed.Firstly,the fusion model of heterogeneous Numbers was established,the redundant data was eliminated through correlation processing,and the method of transforming coordinate systems was used to complete multi-sensor spatial alignment.Secondly,the histogram feature method combined with the two-dimensional Fourier transform was used to improve the contrast of the phase spectrum image,obtain a reasonable grayscale order of magnitude,and complete the target feature extraction.Finally,the conflict coefficient between features was improved,and the features whose conflict coefficient exceeds the threshold value were processed,so as to reduce the influence of high-conflict features on the whole recognition process,and the design of the target recognition algorithm was complete.The simulation results show that the accuracy and standard deviation of the algorithm are better than the traditional method,and the reliability of the algorithm is verified.
作者 金华 郑春 把萍 JIN Hua;ZHENG Chun;BA Ping(Admission Office,Anhui Sanlian College,Hefei 230601,China)
出处 《成都工业学院学报》 2021年第2期47-50,共4页 Journal of Chengdu Technological University
基金 安徽三联学院课题(PTZD2020013)。
关键词 物联网 异构传感数据 目标跟踪识别 直方图特征法 二维傅里叶变换 internet of things heterogeneous sensor data target tracking and identification histogram feature method two-dimensional fourier transform
  • 相关文献

参考文献15

二级参考文献85

共引文献205

同被引文献31

引证文献3

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部