K. W. K. Cheung

Fast reconstruction of sectional images in digital holography

Past research has demonstrated that a three-dimensional object scene can be converted into a digital hologram. Subsequently, the object scene can be reconstructed from the hologram with an iterative blind sectional image reconstruction (BSIR) method. However, the computation is extremely intensive, and escalated with the size of holograms. To overcome this problem, we propose a fast BSIR method that reconstructs sectional images with less out-of-focus haze. While the technique proposed here is applicable in general to holography for sectioning, we use holograms acquired by optical scanning holography as examples to show the methods effectiveness.

Enhancing the pictorial content of digital holograms at 100 frames per second

We report a low complexity, non-iterative method for enhancing the sharpness, brightness, and contrast of the pictorial content that is recorded in a digital hologram, without the need of re-generating the latter from the original object scene. In our proposed method, the hologram is first back-projected to a 2-D virtual diffraction plane (VDP) which is located at close proximity to the original object points. Next the field distribution on the VDP, which shares similar optical properties as the object scene, is enhanced. Subsequently, the processed VDP is expanded into a full hologram. We demonstrate two types of enhancement: a modified histogram equalization to improve the brightness and contrast, and localized high-boost-filtering (LHBF) to increase the sharpness. Experiment results have demonstrated that our proposed method is capable of enhancing a 2048x2048 hologram at a rate of around 100 frames per second. To the best of our knowledge, this is the first time real-time image enhancement is considered in the context of digital holography.

Real-time relighting of digital holograms based on wavefront recording plane method.

Relighting is an important technique in photography which enables the optical properties of a picture to be modified without retaking it again. However, different from an optical image, a digital hologram cannot be relit by simply varying the value of individual pixel, as each of them is representing holistic information of the entire object scene. In this paper, we propose a fast method for the relighting of a digital hologram. First, the latter is projected to a virtual wavefront recording plane (WRP) that is located close to the object scene. Next, the WRP is relit, and subsequently expanded into a full hologram. Experiment results have demonstrated that our proposed method is capable of relighting a 2048x2048 hologram at a rate of over 50 frames per second. To the best of our knowledge, this is the first time relighting is considered in the context of holography.

Low-bit-rate computer-generated color Fresnel holography with compression ratio of over 1600 times using vector quantization [Invited]

We propose a method for compressing a digital color Fresnel hologram based on vector quantization (VQ). The complex color hologram is first separated into three complex holograms, each representing one of the primary colors. Subsequently, each hologram is converted into what we call a real Fresnel hologram and compressed with VQ based on a universal codebook. Experimental evaluation reveals that our scheme is capable of attaining a compression ratio of over 1600 times and still preserving acceptable visual quality on the reconstructed images. Moreover, the decoding process is free from computation and highly resistant to noise contamination on the compressed data.

Binary Mask Programmable Hologram

We report, for the first time, the concept and generation of a novel Fresnel hologram called the digital binary mask programmable hologram (BMPH). A BMPH is comprised of a static, high resolution binary grating that is overlaid with a lower resolution binary mask. The reconstructed image of the BMPH can be programmed to approximate a target image (including both intensity and depth information) by configuring the pattern of the binary mask with a simple genetic algorithm (SGA). As the low resolution binary mask can be realized with less stringent display technology, our method enables the development of simple and economical holographic video display.

Demonstration of compression ratio of over 4000 times for each digital hologram in a sequence of 25 frames in a holographic video

Past research has demonstrated that computer-generated Fresnel holograms can be compressed by over 1600 times based on vector quantization (VQ). In this approach, a digital hologram is first partitioned into non-overlapping square image blocks, each represented as a source vector. Compression is achieved by substituting each source vector with a single index pointing to the nearest member within a small codebook of codevectors. Despite the success, the compression time is rather lengthy, and although the compression ratio can be further increased by enlarging the size of the image block, the quality of the reconstructed image will be degraded significantly. In this paper, we propose a very low bit-rate method for hologram compression based on the integration of VQ, decimation, and the Burrows–Wheeler encoder. Experimental evaluation reveals that our scheme is about 4 times faster than the existing method, and attains higher coding fidelity in terms of quantitative measurement. We also demonstrate an average compression ratio of over 4000 times for each digital hologram in a sequence of 25 frames in a holographic video, maintaining favorable visual quality on the reconstructed images. To our understanding, this is the highest compression ratio that has ever been attained in digital holography.

Fast numerical generation and hybrid encryption of a computer-generated Fresnel holographic video sequence

Research demonstrates that a Fresnel hologram can be generated and simultaneously encrypted numerically based on a secret symmetric key formed by the maximal length sequence (M-sequence). The method can be directly extended to encrypt a video holographic clip in a frame-by-frame manner. However, given the limited combination of signals in the family of M-sequence, hacking the secret key through trial and error can be time consuming but not difficult. In this letter, we propose a method that is difficult to crack with brute force for encrypting a holographic video sequence. An M-sequence is first randomly assigned to encrypt each frame of the holographic video signal. Subsequently, the index of the selected M-sequence, which is necessary to decrypt the hologram, is encrypted with the RSA algorithm before transmitting to the receiving end. At the receiving end, the decoder is provided with a private key to recover the index for each frame, and the corresponding M-sequence is used to decrypt the encoded hologram.