传统的光学投影层析受相机感光元件动态范围及曝光时间的限制, 难以对具有复杂空间结构分布的样品获取其完整且精细的三维结构信息。针对传统光学投影层析三维成像系统存在的问题, 提出了在传统光学投影层析技术中引入朗伯体光源以及线性化动态范围变换的新方法。使用朗伯体光源照射样品, 通过多次曝光分别对样品进行图像采集, 获取相机实际响应曲线, 线性化处理曲线中非线性响应区域以解决传统多次曝光动态范围变换存在的非线性失真和假象问题, 然后应用图像融合技术对多次曝光获取的原始图像数据进行融合, 运用反投影算法重构样品三维成像, 从而获得具有复杂空间结构样品的精细三维结构信息。理论分析与成像结果表明, 这种基于光学投影层析的三维结构成像新方法可以获得复杂空间结构样品更多的信息。

The dynamic range and exposure time of the camera are the main limitation for the traditional optical projection tomography to achieve the intact and detailed three-dimensional information of samples with complicated spatial structure. The normalized dynamic range-transform of three-dimensional imaging acquisition and process are used to solve the existing conundrum of traditional three-dimensional imaging based on the optical projection tomography technology, by means of setting the Lambert source and carrying on the new method of normalized dynamic range-transform. Setting multiple exposures to get the image of sample, normalizing the tested response curve of the camera for solving the non-linear distortion and false appearance existing in the traditional high dynamic range, using the image fusion theory to improve the image processing program, then reconstructing the sample by the reverse projection algorithm, and get the three-dimensional structure of the sample finally. From the analysis of imaging result, the normalized dynamic range-transform of three-dimensional imaging acquires more information of samples with complicated spatial structure, more details of different areas are able to be observed through this way.