基于玫瑰线扫描双色信息融合目标识别方法

Target Recognition Approach Based on the Fusion of Dual Color Rosette Scanning Data

图形扫描复合制导是当今深受国内外重视的一种制导方式。虽然玫瑰线扫描双色导引头技术已在战术地空导弹中得到了应用，但目前主要将其作为一维信号加以处理，不利于目标的识别与信息融合。本文的基本思想在于将玫瑰线扫描得到的一维波形转化为二维准成像图像，同时引入目标、背景以及诱饵的红外与紫外概率模型，采用Ｄｅｍｐｓｔｅｒ证据理论将红外和紫外准成像图像进行融合，以提高系统去除诱饵即假目标的能力。模拟实验结果证明了本文所提出方法的有效性和可行性。

Compound guidance of graphics scanning is heavily concentrated in both inland and overseas nowadays.Although rosette scanning with infrared and violet sensors had been widely applied in missile guidance system,but usually it is treated as an 1 D signal,unsuitable for target recognition and data fusion.The basic idea in this paper is to transform 1 D signal of rosette scanning to a 2 D quasi image and the models of infrared and violet of target,background and false target to be known.Then the Shafer Dempster reasoning is used to fuse infrared and violet quasi images,making the ability of elminating false target of system increase.The simulated experiments demonstrate the efficiency and feasibility of our approach.