|
In this paper, a fast hologram generation method based on optimal segmentation of sub-computer-generated hologram is proposed. When using the point light source (PLS) model for computer-generated hologram (CGH), the 3D object is discretized into individual object points, which are assumed to independent ideal sources. Then, the sub-CGH can be obtained by calculating the diffraction light field distribution of each ideal PLS, and the final hologram can be generated by further calculating of the sub-CGHs. In the proposed method, the contribution of any pixel in the sub-CGH generated by each ideal PLS to the final reconstructed image is calculated, and each sub-CGH is divided into the optimized diffraction areas (ODAs) and the invalid diffraction areas. Then, only the ODAs of each sub-CGH is pre-calculated and saved for further generating the final hologram. Because of the size of the ODAs are much smaller than that of the sub- CGHs, the hologram generation speed is greatly improved. The proposed method can be used to optimize most of the PLS model-based hologram generation methods. The proposed method is used to optimize the novel look-at-table (NLUT) method and wavefront recording plane (WRP) method, and both of them achieve good acceleration effect. In the optimization of the NLUT method, the sub-CGH database of the traditional NLUT method is replaced by the ODAs database. The final hologram can be generated by calling and superimposing all the ODAs. In the optimization of the WRP method, the WRP is calculated by using the ODAs instead of the sub-CGHs. Then, the final hologram is generated by using the Fresnel diffraction from the WRP to the holographic plane. With the proposed method, the calculation speed of the NLUT method and the WRP method can be significantly improved and the quality of the reconstructed image is not affected.
|
