赝像-正像转换算法在全息体视图打印中的应用

Application of Pseudoscopic-Orthoscopic Conversion Algorithm in Holographic Stereogram Printing

将适用于三维场景信息非对称采集和显示的赝像-正像转换(POC)算法应用于全息体视图打印,得到了采样平面与全息图平面的距离、采样间隔与全息单元尺寸比例不同时,采样图像和合成视差图像之间的像素匹配关系。分析了曝光光学系统参数和POC算法参数对体视图再现像视场角的影响,得到了场景深度和视场角之间的制约关系。实验通过在不同场景深度下三维物体的再现,验证了POC算法对全息体视图打印的适用性以及场景深度和视场角之间制约关系的正确性。实验结果表明,当场景深度较小时,再现像重影导致分辨率降低,并从理论上分析了全息体视图重影现象产生的原因。像素级精确的视差图像避免了数据误差对体视图再现质量的影响,对提高再现像分辨率有积极意义。

The pseudoscopic-orthoscopic conversion (POC) algorithm, which is applicable to asymmetrical capture and display of three-dimensional (3D) scene information, is utilized to perform the holographic stereogram printing. The matching relationship between pixels on the sampled image and the synthetic parallax image is obtained under different distances between the sampling plane and the holographic plane and different ratios of sampling interval to holographic unit size. The influences of exposure optical system parameters and POC algorithm parameters on the field of view of the stereogram are also analyzed, and the relationship between scene depth and field of view is obtained. The experimental results demonstrate the applicability of POC algorithm to holographic stereogram printing and the validity of the relationship between scene depth and field of view by reconstructing 3D objects at different scene depths. The resolution is reduced because of the reconstructed image ghosting when the depth of the scene is small, and reasons for the ghosting of the holographic stereogram are also explained. The accurate parallax image at the pixel level avoids the influence of data error on the quality of holographic stereogram, and has positive significance in improving the resolution of the reconstructed image.