基于奇值分解的三维光学切片显微图像恢复算法研究

Image restoration algorithms based on singular value decomposition of 3-D optical sectioning microscopy

阐述了光学切片显微技术的基本原理 ,分析了显微镜光学系统点扩展函数循环矩阵的奇异性造成荧光图像恢复质量大大降低所导致的病态问题 ,应用降质模糊矩阵的奇值分解方法 ,挑选出较大特征值 ,并将空域问题的求解通过傅立叶变换转到频率域 ,使计算的复杂度降低 ,最终得到较为理想的复原图像。

The paper discusses the basic theory of optical sectioning microscopy, and also analyzes the ill posed problem caused by the degradation in fluorescence images resulting from the singularity of point spread function circular matrix of the microscope system. By applying the singular value decomposition, the bigger eigenvalues are selected and the ill posed problem is solved. With the fast Fourier transform, this problem is transformed from space to frequency domain, which simplifies the computation process. Finally, a good result is achieved.