基于改进Hough变换的检测前跟踪算法  

TBD Algorithm Based on Improved Hough Transform

作　　者：李晓聪[1] 涂刚毅[1] 裴江[1] 吴少鹏[1] 

机构地区：[1]中国船舶重工集团公司第七二四研究所,江苏南京211106

出　　处：《现代防御技术》2016年第5期137-142,共6页Modern Defence Technology

基　　金：某预先研究项目

摘　　要：针对密集杂波环境中的弱小目标检测问题,提出了一种改进Hough变换的检测前跟踪算法,解决Hough变换参数空间峰值簇拥导致虚假轨迹较多的问题。采用样本方差信息对非相参积累值进行加权,利用时间信息抑制部分杂波和噪声积累峰值,改进非相参积累的积累方法;依据各单元的积累值对局部峰值簇拥的几个单元进行融合处理,改进阈值提取的峰值提取方法。仿真分析表明,改进算法检测目标虚假轨迹少于标准Hough变换方法,与先检测后跟踪方法相比,改进算法有3~4 d B信噪比得益。In view of the detection of the dim targets in dense clutter environment,a TBD algorithm based on improved Hough transform is proposed to solve the increasing false tracks caused by the peak clustering in parameter space. The non-coherent integration values are weighed by the samples＇ variance and modified by the time information in this new accumulation method. The improved peak extraction method calculates the fusion parameters of several cells surrounding the local peak. The simulation result shows that the improved algorithm has fewer false tracks than the standard Hough transform method and has the benefit of 3～4 d B signal to noise ratio compared with the traditional method.

关 键 词：HOUGH变换 检测前跟踪 非相参积累 弱小目标检测 先检测后跟踪 峰值提取 

分 类 号：TN957.51[电子电信—信号与信息处理] TP301.6[电子电信—信息与通信工程]