1种机场场面移动目标特征提取方法

A Method of Moving Object Feature Extraction at Airport

目标特征提取作为其类型识别的关键问题,对保证机场场面运行安全有着重要的研究意义。基于地磁传感器(anisotropic magneto resistance,AMR)采集的机场移动目标磁信号,提出了1种综合考虑磁信号的时域及频域特征的目标检测方法,内容包括目标检测,磁信号的时域及频域特征提取。该方法首先通过在跑道中线布置双传感器方式感知信号并计算目标长度,基于目标长度规整时域信号获得其峰值特征;其次利用小波变换3层分解已规整的时域信号,并归一化各阶及近似信号能量值作为提取信号的频域特征,最终与信号时域特征共同构成信号的全部特征。在实验验证中,利用目标特征区分度测度方法使得不同目标之间的相似度均低于0.79,相比于单一时域或频域特征提取方法,时频特征提取方法消除了不同目标间极端相似的现象。

Object feature extraction was the key to target identification and has great significance to ensure the safe operation of airport.A method of combining time domain feature extraction and frequency domain feature extraction based on Anisotropic Magneto Resistance was proposed.This method included object detection,time domain feature extraction and frequency domain feature extraction.Firstly,two sensors were installed under center line of the runway to sense magnetic signals and calculate object lengths.Peak characteristics of signals were obtained based on regulating the time domain signal according to the object.Secondly,the regulated signals were decomposed into 3layers wavelet coefficients,with normalizing approximate signal energy and every layer signal energy as frequency domain features of the signals.At last,all features of a signal consisted of time domain features and frequency domain features.The similarities of different objects were restricted below 0.79 using the method of target features difference measurement during the experimental validation.The results indicated the identification from the proposed method avoided the high similarity between different objects comparing the method of time domain feature extraction or frequency domain feature extraction only.