用多项式近似逆滤波函数实现场曲修正

根据光电成像系统传递函数,确定逆滤波函数对获得图像进行反降质恢复.将连续逆滤波函数按泰勒级数展开,用多项式近似表示.对逆滤波函数的多项式作反傅里叶变换,得到图像恢复在空间域中的近似运算表达,即图像信号及其各阶导数的线性组合,而不是复杂的反卷积操作.该方法特别适合空间移变系统的图像恢复.详细分析了这种方法的原理,推导了运算公式,并给出了对场曲图像处理的结果.

Correction of Field Curved Images by the Polynomial Approximation of the Inverse Filtering Function

When an optical transfer function is known,the inverse filtering function can be determined and used for image recovery. If the inverse filtering function is continuous, it can actually be represented as a Taylor polynomial. Hence the recovery of images may be approximately realized in spatial domain by the linear combination of the image and its derivatives rather than by the complex deconvolution. It is considerable for the image recovery of the spatially variant system, such as the correction of field curved images, as the spatial variance only reflects in the coefficients of the linear combination. The recovery equations are derived and the processing results are presented.