Assessment of 3D reconstructed image quality for computer-generated holographic 3D display

A methodology of assessing a three-dimensional (3D) reconstructed image quality objectively for computergenerated holographic 3D display based on evaluating the reconstructed optical field is proposed. According to the concept of ‘holographic view-window’ (HVW), the point source method is used to calculate the ‘ideal’ Fresnel hologram and the precise diffraction field is numerically simulated after the 3D model is finely sampled. This accurate diffraction field is used as the evaluation criteria that can be regarded as ‘ground truth’. In comparison with the reconstructed images of the 3D diffraction field obtained by other common algorithms, such as point source method (PSM), Gerchberg-Saxton iterative (GS) algorithm and frequency-filtering method (FFM), the rationality of the evaluation function—the peak signal-to-noise ratio (PSNR) for the image quality assessment (IQA) of the 3D reconstructed diffraction field is investigated comprehensively at different reconstructed distances. The simulation results indicate that the PSNR is relatively reasonable for the evaluation of 3D reconstructed images taking into account the defocusing phenomena. The investigation would suggest an alternative optical scheme for the objective assessment of 3D image quality which provides a theoretical basis for the detection of holographic 3D image quality, the improvement of holographic 3D display algorithms and the design of the future spatial light modulator (SLM)