基于自适应形态学边缘检测的时频检测算法  

TIME-FREQUENCY DETECTION ALGORITHM BASED ON ADAPTIVE MORPHOLOGICAL EDGE DETECTION

作　　者：赵方超[1] 蒋建中[1] 郭军利[1] 朱世先 

机构地区：[1]解放军信息工程大学信息系统工程学院 [2]解放军91746部队

出　　处：《计算机应用与软件》2014年第9期279-282,共4页Computer Applications and Software

基　　金：国家高技术研究发展计划项目(2009ZX03006-008)

摘　　要：针对通信对抗中跳频信号的检测问题,提出一种基于自适应形态学边缘检测的时频检测算法。该算法把跳频信号的时频谱图SP(Spectrogram)视作一幅图像,首先利用多尺度结构元素对其滤波,再根据图像边缘方向自动选择相应方向的结构元素进行边缘检测,最后通过与设定的阈值比较判决跳频信号是否存在。仿真结果表明,此算法可以有效去除噪声和干扰信号,增强跳频检测系统鲁棒性、提高检测效率。Aiming at frequency-hopping （FH）signals detection in communication confrontation,a time-frequency detection algorithmbased on adaptive morphological edge detection is proposed.In the algorithm,the time-frequency spectrogram of FH signals is regarded as theimage,and the multi-scale structure element is used to filter its noise first.Then,the structure element in corresponding direction isautomatically selected according to the direction of the image edge to detect the edges.Finally,by comparing with the threshold value set,whether there are the FH signals or not is determined.Simulation result demonstrates that the algorithm can effectively remove the noise andinterference signals,enhance the robustness of FH detection system,and improve the detection efficiency.

关 键 词：跳频检测 时频分布 形态学滤波 自适应边缘检测 

分 类 号：TN911[电子电信—通信与信息系统]