Multiple-image encryption based on triple interferences for flexibly decrypting high-quality images.

We propose a multiple-image encryption (MIE) scheme based on triple interferences for flexibly decrypting high-quality images. Each image is discretionarily deciphered without decrypting a series of other images earlier. Since it does not involve any cascaded encryption orders, the image can be decrypted flexibly by using the novel method. Computer simulation demonstrated that the proposed method's running time is less than approximately 1/4 that of the previous similar MIE method. Moreover, the decrypted image is perfectly correlated with the original image, and due to many phase functions serving as decryption keys, this method is more secure and robust.

Generation speed and reconstructed image quality enhancement of a long-depth object using double wavefront recording planes and a GPU.

A method for fast computer hologram generation for long-depth objects using double wavefront recording planes (WRPs) and a graphics-processing unit (GPU) is presented. The WRPs are placed between the object and the hologram plane. Each WRP records the wavefront from a section of the object. Double WRPs can provide a shorter calculation time and enhanced reconstructed image quality compared with a single WRP, especially for long-depth objects. The average generation speed of two WRPs is 2.5 times that of one WRP. The correlation efficiency of the reconstructed layer relative to the original is 94% for two WRPs and 88.3% for one WRP at the close depth layer.

Synthesis of computer-generated spherical hologram of real object with 360° field of view using a depth camera.

A method for synthesizing a 360° computer-generated spherical hologram of real-existing objects is proposed. The whole three-dimensional (3-D) information of a real object is extracted by using a depth camera to capture multiple sides of the object. The point cloud sets which are obtained from corresponding sides of the object surface are brought into a common coordinate system by point cloud registration process. The modeled 3-D point cloud is then processed by hidden point removal method in order to identify visible point set for each spherical hologram point. The hologram on the spherical surface is finally synthesized by accumulating spherical waves from visible object points. By reconstructing partial region of the calculated spherical hologram, the corresponding view of the 3-D real object is obtained. The principle is verified via optical capturing using a depth camera and numerical reconstructions.

Error analysis in parallel two-step phase-shifting method.

This paper presents an analysis of error in parallel two-step phase-shifting method. From the analysis, it is clarified that the maximum bandwidth of the object that this technique can capture is a half bandwidth of recording devices. Also, the recording distance must be two times longer than the conventional method to have a good reconstruction image. The analysis was verified by simulations and experimental results.