基于FFT的三维扫描图像深层损耗快速恢复方案

An FFT Based Restoration Algorithm for Deep Layer Attenuation of Three dimensional Scanning Images

在分析三维共焦扫描图像衰减数学模型的基础上，针对其深层损耗进行了不依赖于试样的恢复。在确保恢复准确性的前提下推导出基于ＦＦＴ（快速傅里叶变换）的快速算法，该算法具有准确、高效、适应性强、不引入额外噪声等优点。讨论了算法的计算量、内存占用量及其在微机上的实现。提出相应的改善措施，并给出测试结果。

A “sample independence” restoration method aiming at deep layer attenuation correction of three dimensional confocal scanning images is given with the analysis of its mathematical models presented in details. A rapid access based on FFT is derived with the presumosition that the resulted images will be precise enough. The algorithm, without additional noise inducement, is also proved to be accurate, efficient and flexible. The computing efficiency and memory consumption are discussed followed by its computational realization. Approaches for better performance are proposed subsequently. The successful applications to both test images and real ones are also given.