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。 Based on the analysis of the mathematic model of 3D confocal scanning image attenuation, it does not depend on the sample recovery for its deep loss. The fast algorithm based on FFT (Fast Fourier Transform) is derived under the premise of ensuring the accuracy of recovery. The algorithm has the advantages of accuracy, high efficiency, adaptability and no additional noise. The calculation of the algorithm, memory consumption and its implementation on the computer are discussed. Put forward the corresponding improvement measures, and give the test results.