作　　者： ; ; ;

机构地区： 深圳大学信息工程学院

出　　处： 《计算机工程与应用》 2010年第30期26-29,共4页

摘　　要： 针对复杂环境下红外图像信噪比和对比度低,边缘模糊,目标分割困难的情况,提出一种基于模糊增强和均值漂移图像滤波的红外目标分割方法。首先定义新的隶属度函数,运用模糊集理论进行红外图像增强,避免了传统模糊增强算法的弊病,有效提高目标与背景的对比度;之后利用IC(I交叉置信区)规则确定均值漂移的带宽参数,提出一种新的自适应带宽均值漂移图像滤波方法,实现图像的进一步平滑和聚类;最后利用自适应阈值实现红外目标分割。实验结果表明,算法能够正确有效地分割出复杂环境下的红外目标,并且很好地保持了目标的轮廓细节。 Infrared images always have low SNR and contrast,and boundaries of infrared target are blurry.Therefore the segmentation of infrared target in complex environment is very difficult.A new algorithm for infrared target segmentation based on fuzzy enhancement and mean shift is proposed in this paper.Firstly,a new membership function of fuzzy is defined,the contrast between target and background in infrared image is improved efficiently by the enhancement method based on the fuzzy set theory,in which the disadvantages of traditional fuzzy enhancement methods are avoided.Then,the intersection of confidence intervals（ICI） rule is used to determine the bandwidth in mean shift,and a new adaptive bandwidth mean shift algorithm is presented to realize further smoothing and clustering of image.Finally,infrared target is segmented by the adaptive threshold.Experimental results indicate that the algorithm can segment the infrared target under complex environment correctly and efficiently,and the good details of target are reserved.

关 键 词： 模糊集 图像增强 均值漂移 红外目标 目标分割

领　　域： [自动化与计算机技术] [自动化与计算机技术]