Optimized method of complex-amplitude holographic display and computer-generated holograms (CGHs) based on complex-valued convolutional neural network are presented. The wavefront frequency decomposition is carried out on the object light wave to generate two frequency images. The low frequency image is determined by amplitude CGH loaded with an amplitude spatial light modulator (SLM). The high frequency image is decided by phase CGH loaded with a phase SLM. The mentioned two sub-holograms are reconstructed and integrated at the identical location, and the amplitude and phase information are enhanced. Complex-valued convolutional neural network (CCNN) is used for real-time CGH generation. Complex-valued convolutional and ReLU layers are combined in CCNN, which is designed based on the feature of complex amplitude. Instead of processing amplitude and phase respectively, CCNN processes the complex amplitude directly. CCNN-CGH architecture can generate 1920×1072 and 3840×2160 resolution CGH with an average PSNR of more than 30dB in real-time. The speckle noise in display images is reduced, which gives the credit to the reconstruction of complex-amplitude modulation. The experimental results show that the method simultaneously enables high image quality and efficient calculation to be provided, which is more suitable for the real time 3D display. The quality of reconstructed images can be noticeably improved.
Holographic display can reconstruct the optical wave field of three-dimensional (3D) scenes. Computer- generated hologram (CGH) can simulate the optical transmission process from the 3D object to the hologram plane and avoid complicated experiments. Wavefront recording plane (WRP) is a fast algorithm for CGH, which is placed between the 3D object and the CGH. However, the computational amount is increased when the depth range of the 3D object is expanded. In this paper, we propose a fast calculation algorithm for real-time interactive CGH using a novel Fresnel zone plate (FZP) extension method and ray tracing. In the first step, ray tracing is used to get the amplitude value and the depth value of each object point. The maximum resolution of each FZP is determined according to the distance from the object point to the WRP. In the second step, a preset square portion of each FZP is used to calculate the complex amplitude of the WRP. Then diffraction calculation is performed from the WRP to the hologram plane for calculating CGH, and the computational domain of each FZP is gradually extended as the number of frames increases. The above calculation is repeated when interacting. The computational complexity is effectively reduced because the resolution of each FZP is increased until the maximum resolution is reached, instead of directly calculating each FZP with the maximum resolution. Experiments demonstrate the proposed method can generate CGH in real-time. The resolution of the hologram is 1024x024 and the CGH generation rate is 45.81 frames per second when interacting.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.