根据成像偏振探测中stokes参数图像获取的原理和方法,分析了传统求取stokes参数图像方法在保留图像细节信息和处理输入图像微小平移等方面的不足,在此基础上提出了基于自适应平移不变小波分解的stokes参数图像求取方法,并用无下采样的提升框架来执行平移不变小波分解以提高算法的灵活性和计算速度.针对不同的stokes参数图像求取方法,用均值、方差、熵、边界能量对融合结果进行评价和分析.通过仿真,表明本文方法提高了stokes参数图像的质量并改善了stokes参数图像计算方法对输入图像平移的敏感性.

According to the principle and method of Stokes images′ calculation in imaging polarization detection,the shortcomings of traditional method in preserving detail information and dealing with the shifting of scene are analyzed.Then a new calculation method is proposed based on adaptive shift invariant wavelet. To make the algorithm more flexible and fast,lifting scheme is used to perform wavelet transformation.The average,variance,entropy,and boundary energy are used to evaluate the performance of different calculation methods.Through simulation,the quality of Stokes′ images are improved by this paper′s algorithm,at the same time,this algorithm is insensitivity to shifting of scene in source image.