Myungjin Cho

Integral imaging is a promising technology for passive three-dimensional (3D) scene acquisition and display, providing full parallax and continuous viewpoints without the need for specialized viewing glasses. However, in volumetric computational reconstruction (VCR), depth estimation accuracy is basically constrained by shifting pixel quantization errors arising from the discrete pixelation of elemental images. Although non-uniform VCR (NVCR) can reduce these errors through localized rounding, shifting pixel quantization errors may not be reduced to exactly zero. In this paper, we propose a computational reconstruction method for integral imaging that completely eliminates shifting pixel quantization errors while preserving uniform depth resolution across multiple target depths. The proposed method uses a least common multiple (LCM) search based on the respective optimum pitch bases of multiple target planes to calculate a unified optimum pitch. Simultaneously, the physical layout of the camera array is structured non-uniformly using a Latin-square distribution of subarrays governed by the approximate squared-multiple rule. This layout maintains a constant depth resolution limit across the entire distance range and avoids the quadratic degradation of depth resolution observed at farther distances. Through mathematical analysis and computational simulations, we demonstrate that the proposed non-uniform pitch method achieves exact depth recovery with a mean squared error of zero (MSE = 0) at all target depths and produces sharp, high-intensity peak sidelobe ratio (PSR) responses. This approach holds significant potential for advanced sensing applications, such as autonomous vehicles, digital twins, and industrial robotics.

A notable drawback of Double Random Phase Encryption (DRPE), a prominent optical cryptography technique, is its low data efficiency. This is because both the encrypted image and the decryption key are represented as complex numbers. To address this issue, a conventional method was proposed that encrypts two images simultaneously by treating the first image as amplitude and the second image as phase. Nevertheless, processes such as integral imaging, which extract 3D object information from images, utilize vast amounts of imagery, necessitating further enhancements in data efficiency. The objective of this research is to enhance DRPE and improve data efficiency by increasing the number of images that can be processed simultaneously. This paper incorporates the information from a third image into the random phase mask used in conventional methods, enabling the simultaneous processing of three images. It also proposes a method to synthesize two images by extracting their high-order bits and combining them. The combination of this image composition method as a preprocessing step with the proposed DRPE method enables the simultaneous processing of six images. As a result, the proposed method achieves a data efficiency approximately six times that of the basic DRPE and approximately three times that of conventional methods. The quality of the decrypted images was evaluated using PSNR and SSIM, while the encryption strength was assessed in terms of key space, key sensitivity, entropy, and correlation coefficients.

In fire and emergency rescue operations, visibility is often severely degraded by smoke, airborne debris, or atmospheric pollutants including smog and yellow dust. Several image restoration techniques, including Dark Channel Prior (DCP), Color Attribution Prior (CAP), Peplography, and Adaptive Removal via Mask for Scatter (ARMS), have been proposed to recover clear images under such conditions. However, these methods exhibit significant limitations in heavy scattering environments. This paper proposes a novel visibility restoration method for disaster situations, building upon the state-of-the-art ARMS method. To maximize the suppression of scattering effects, the Scattering Media Model is refined through Gaussian estimation. Additionally, an overlapping matrix is introduced to effectively handle non-uniformly distributed scattering conditions. The proposed method is evaluated using a real rescue operation image dataset provided by the Fire and Disaster Management Agency of Japan. Qualitative visual assessments and quantitative performance metrics demonstrate that the proposed approach significantly outperforms conventional methods under severe scattering conditions.

In this paper, we propose a compressionmethod for optical encryption using singular value decomposition (SVD). Double random phase encryption (DRPE), which employs two distinct random phase masks, is adopted as the optical encryption technique. Since the encrypted data in DRPE have the same size as the input data and consists of complex values, a compression technique is required to improve data efficiency. To address this issue, we introduce SVD as a compression method. SVD decomposes any matrix into simpler components, such as a unitary matrix, a rectangular diagonal matrix, and a complex unitary matrix. By leveraging this property, the encrypted data generated by DRPE can be effectively compressed. However, this compression may lead to some loss of information in the decrypted data. To mitigate this loss, we employ volumetric computational reconstruction based on integral imaging. As a result, the proposed method enhances the visual quality, compression ratio, and security of DRPE simultaneously. To validate the effectiveness of the proposed method, we conduct both computer simulations and optical experiments. The performance is evaluated quantitatively using peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and peak sidelobe ratio (PSR) as evaluation metrics.

Optical metasurfaces, ultrathin 2D artificial materials with subwavelength structures, offer exceptional electromagnetic manipulation capabilities, making them promising for photonic integration. Traditional design methods rely on the local periodic approximation (LPA), treating unit cells as infinite periodic arrays to reduce computational cost. However, this approach neglects near-field coupling between adjacent units, resulting in focal shifts of 9%–22% in metalenses. We propose an adjoint method-based optimization framework that accounts for inter-unit coupling via full-wave simulations, creating a parametric gradient optimization space to refine metalens designs. Experimental validation through simulations demonstrates that the optimized metalens achieves a focal shift of 0.1%–1.8%, significantly outperforming LPA-based designs while showcasing scalable optimization potential. This framework offers a novel, efficient approach for metasurface optimization, with applications in optical displays, augmented reality, and 3D imaging.

Digital Holographic Microscopy (DHM) is a method of converting hologram images into three-dimensional (3D) images by image processing, which enables us to obtain the detailed shapes of the objects to be observed. Three-dimensional imaging of the microscopic objects by DHM can contribute to the early diagnosis and the detection of the diseases in the medical field by observing the shape of the cells. DHM requires several experimental components. One of them is the laser, which is a problem because its high power may cause the deformation and the destruction of the cells and the death of the microorganisms. Since the greatest advantage of DHM is the detailed geometrical information of the object by 3D measurement, the loss of such information is a serious problem. To solve this problem, a Neutral Density (ND) filter has been used to reduce power after the laser irradiation. However, the image acquired by the image sensor becomes too dark to obtain sufficient information, and the effect of noise increased due to the decrease in the amount of light. Therefore, in this paper, we propose the Frame-Stacking Method (FSM) for dark DHM for reproducing 3D profiles that enable us to observe the shape of the objects from the images taken in low-power environments when the power is reduced. The proposed method realizes highly accurate 3D profiles by the frame decomposition of the low-power videos into images and superimposing and rescaling the obtained low-power images. On the other hand, the continuous irradiation of the laser beam for a long period may destroy the shape of the cells and the death of the microorganisms. Therefore, we conducted experiments to investigate the relationship between the number of superimposed images corresponding to the irradiation time and the 3D profile, as well as the characteristics of the power and the 3D profile.

In this paper, we propose a new method for three-dimensional (3D) visualization that proportionally estimates the number of photons in the background and the object under photon-starved conditions. Photon-counting integral imaging is one of the techniques for 3D image visualization under photon-starved conditions. However, conventional photon-counting integral imaging has the problem that a random noise is generated in the background of the image by estimating the same number of photons in entire areas of images. On the other hand, our proposed method reduces the random noise by estimating the proportional number of photons in the background and the object. In addition, the spatial overlaps have been applied to the space where photons overlap to obtain the enhanced 3D images. To demonstrate the feasibility of our proposed method, we conducted optical experiments and calculated the performance metrics such as normalized cross-correlation, peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM). For SSIM of 3D visualization results by our proposed method and conventional method, our proposed method achieves about 3.42 times higher SSIM than conventional method. Therefore, our proposed method can obtain better 3D visualization of objects than conventional photon-counting integral imaging methods under photon-starved conditions.

In this paper, we propose automation of estimating scattering media information in peplography using singular value decomposition (SVD) and discrete cosine transform (DCT). Conventional scattering media-removal methods reduce light scattering in images utilizing a variety of image-processing techniques and machine learning algorithms. However, under conditions of heavy scattering media, they may not clearly visualize the object information. Peplography has been proposed as a solution to this problem. Peplography is capable of visualizing the object information by estimating the scattering media information and detecting the ballistic photons from heavy scattering media. Following that, 3D information can be obtained by integral imaging. However, it is difficult to apply this method to real-world situations since the process of scattering media estimation in peplography is not automated. To overcome this problem, we use automatic scattering media-estimation methods using SVD and DCT. They can estimate the scattering media information automatically by truncating the singular value matrix and Gaussian low-pass filter in the frequency domain. To evaluate our proposed method, we implement the experiment with two different conditions and compare the result image with the conventional method using metrics such as structural similarity (SSIM), feature similarity (FSIMc), gradient magnitude similarity deviation (GMSD), and learned perceptual image path similarity (LPIPS).

Digital Holographic Microscopy (DHM) is a technique that uses the phase information of light to generate a three-dimensional (3D) profile of an object. Recently, it has been utilized in various fields such as disease diagnosis and research on microorganisms. In the process in DHM, a narrow region around one of the sidebands from the frequency domain is windowed to avoid noise caused by the direct current (DC) term. However, it may not obtain the high-frequency information about the object. On the other hand, windowing a wide region increases the noise caused by the DC term, and generates the noise in the 3D profile. To solve this trade-off, we propose a noise reduction method using Kalman filter. From the recorded hologram image, we can create the frequency domain. It obtains multiple windowed sidebands centered on multiple pixels at random from the frequency domain. This creates a group of data in which noise is generated randomly. This is regarded as frequency series data, and Kalman filtering is performed. This method can reduce the noise caused by the DC term while acquiring high-frequency information. In addition, this method has the advantage that only one image is needed for frequency series data in the Kalman filter. The effectiveness of the proposed method is verified by comparison with conventional filtering methods and general image processing methods. The validation results prove the usefulness of the proposed method, and the proposed method is expected to have a significant effect on improving the accuracy of disease diagnosis techniques using DHM.