一种亮暗小目标自适应检测方法  

Method for bright and dark small targets adaptive detection

作　　者：王岳环[1] 周晓玮[1] 张天序[1] 

机构地区：[1]华中科技大学图像识别与人工智能研究所多谱信息处理技术国防科技重点实验室,武汉430074

出　　处：《计算机应用研究》2007年第11期289-291,共3页Application Research of Computers

基　　金：国家自然科学基金重点资助项目(60135020);国防预研基金资助项目(51401020203JW0501)

摘　　要：在原始多级滤波检测小目标的方法基础上,改进了多级滤波器的结构,对其差分后的信号同时保留亮暗部分的信息。通过对感兴趣区域进行基于侧抑制的局对比度增强处理,有效地抑制了背景杂波,完整地增强了亮、暗目标,并且对目标边缘增强的定位准确,有利于目标特征的计算。实验结果表明,该算法能对亮、暗目标自适应检测,并且具有良好的检测性能。To solve the problems of detecting bright and dark targets, an improved multi-level filter was proposed based on the method for small targets detection with traditional multi-level filter in order to preserve both bright and dark region information. Through the processing of enhancement of local contrast based on lateral inhibition for attention region, the whole bright and dark targets were enhanced with the clutter inhibited. Meanwhile the edge of the target was enhanced precisely, which was easy to compute the feature of the target. Experimental results demonstrate that the algorithm proposed can detect bright and dark small targets adaptively and effectively.

关 键 词：亮暗小目标 自适应检测 侧抑制 局部对比度 红外 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]