The vacuum degree is the key parameter reflecting the quality and performance of vacuum glass. This investigation proposed a novel method, based on digital holography, to detect the vacuum degree of vacuum glass. The detection system was composed of an optical pressure sensor, a Mach–Zehnder interferometer and software. The results showed that the deformation of monocrystalline silicon film in an optical pressure sensor could respond to the attenuation of the vacuum degree of vacuum glass. Using 239 groups of experimental data, pressure differences were shown to have a good linear relationship with the optical pressure sensor’s deformations; pressure differences were linearly fitted to obtain the numerical relationship between pressure difference and deformation and to calculate the vacuum degree of the vacuum glass. Measuring the vacuum degree of vacuum glass under three different conditions proved that the digital holographic detection system could measure the vacuum degree of vacuum glass quickly and accurately. The optical pressure sensor’s deformation measuring range was less than 4.5 μm, the measuring range of the corresponding pressure difference was less than 2600 pa, and the measuring accuracy’s order of magnitude was 10 pa. This method has potential market applications.

The speckle noise generated during digital holographic interferometry (DHI) is unavoidable and difficult to eliminate, thus reducing its accuracy. We propose a self-supervised deep-learning speckle denoising method using a cycle-consistent generative adversarial network to mitigate the effect of speckle noise. The proposed method integrates a 4-f optical speckle noise simulation module with a parameter generator. In addition, it uses an unpaired dataset for training to overcome the difficulty in obtaining noise-free images and paired data from experiments. The proposed method was tested on both simulated and experimental data, with results showing a 6.9% performance improvement compared with a conventional method and a 2.6% performance improvement compared with unsupervised deep learning in terms of the peak signal-to-noise ratio. Thus, the proposed method exhibits superior denoising performance and potential for DHI, being particularly suitable for processing large datasets.

This study presents a color digital holography (CDH) detection system that includes the optical path and a numerical calculation program to detect tooth erosion under different acid-etching conditions. Etched cattle enamel samples were pressurized within the elastic interval, and the CDH detection system was used to rebuild the three-dimensional deformation field of these samples. Young’s modulus of these samples was employed to evaluate the mechanical properties of enamels by statistical methods. The results show that Young’s modulus of enamel decreases with the increase in etching time, while it decreases slightly with the increase in the etching temperature, and the etching temperature has no effect on the non-etched enamels. Moreover, the detection process proves the CDH detection system has the advantages of high precision, full-field detection, and being non-destructive, fast, and stable. It is expected to have broad application potential for CDH detection systems in the detection of tooth erosion.

The speckle noise generated during digital holographic interferometry (DHI) is unavoidable and difficult to eliminate, thus reducing its accuracy. We propose a self-supervised deep learning speckle denoising method using a cycle-consistent generative adversarial network to mitigate the effect of speckle noise. The proposed method integrates a 4-f optical speckle noise simulation module with a parameter generator. In addition, it uses an unpaired dataset for training to overcome the difficulty in obtaining noise-free images and paired data from experiments. The proposed method was tested on both simulated and experimental data, with results showing a 6.9% performance improvement compared with a conventional method and a 2.6% performance improvement compared with unsupervised deep learning in terms of the peak signal-to-noise ratio. Thus, the proposed method exhibits superior denoising performance and potential for DHI, being particularly suitable for processing large datasets.

The speckle noise generated during digital holographic interferometry (DHI) is unavoidable and difficult to eliminate, thus reducing its accuracy. We propose a self-supervised deep learning speckle denoising method using a cycle-consistent generative adversarial network to mitigate the effect of speckle noise. The proposed method integrates a 4-f optical speckle noise simulation module with a parameter generator. In addition, it uses an unpaired dataset for training to overcome the difficulty in obtaining noise-free images and paired data from experiments. The proposed method was tested on both simulated and experimental data, with results showing a 6.9% performance improvement compared with a conventional method and a 2.6% performance improvement compared with unsupervised deep learning in terms of the peak signal-to-noise ratio. Thus, the proposed method exhibits superior denoising performance and potential for DHI, being particularly suitable for processing large datasets.