Segmented mirror technology plays a pivotal role in the development of future optical telescopes, and achieving co-phasing between the individual mirrors is a critical prerequisite for the successful implementation of this technology. This paper presents a deep learning-based approach for the high-precision detection of fine co-phase errors in segmented mirrors. The proposed method requires only a single far-field image to achieve accurate detection of co-phase errors, demonstrating excellent generalization capabilities. It maintains high detection accuracy even under varying levels of noise and wavefront aberrations. In multi-mirror configurations, the model can be readily adapted to simultaneously detect multiple piston errors through minor structural adjustments. Experimental validation indicates that the root mean square error (RMSE) of the method is 2.9 nm, highlighting its significant potential for practical engineering applications.

The polarization spectroscopic measurement technology can simultaneously measure the polarization parameters and spectral information of light, and it has been widely applied in fields such as astrophysical research and atmospheric remote sensing. Spatially static polarization modulation interferometric spectroscopy combines polarization intensity modulation technology and spatial heterodyne spectroscopy, enabling static and synchronized measurements of high spectral resolution and continuous-band polarization spectral information. In this paper, an experimental setup based on the principles of spatially static polarization modulation interferometric spectroscopy was constructed. Spectral and radiometric calibration of the system were conducted, followed by performance testing to verify the system's spectral and polarization detection capabilities. The experimental results show that the proposed spatially static polarization modulation interferometric spectrometer can achieve a spectral resolution of 7.09 cm-1 and a polarization measurement accuracy better than 0.01.

The structural deformations induced by rocket launch vibrations, on-orbit thermal gradients, and gravitation fluctuations can lead to significant deployment errors for large-aperture, segmented space telescopes. As the size and number of segments increase in future telescopes, the optical-based methods for detecting deployment errors suffer from the range limitations of the millimeter scale and time-consuming processes of the month scale. To address this, we propose a new method for rapid-deployment error detection based on long-range, high-precision capacitive edge sensors. These sensors feature a measurement range of ±13 mm, with a precision better than 7.3 nm, enabling efficient and simultaneous error detection across all segments. This approach significantly reduces the time and steps required compared to traditional optical methods. Through experimental validation, the designed system demonstrated the ability to detect and correct large deployment errors and maintain co-phasing precision, meeting the stringent requirements for future space telescopes. The proposed sensor system enhances deployment efficiency, offering a viable solution for the next generation of segmented space telescopes.

To address the challenges of piston error detection in segmented mirror systems under noisy environments, this article proposes a wavelet-driven deep multichannel feature extraction method (WTCNN). To the best of our knowledge, this is the first application of a multichannel convolutional neural network (CNN) method combined with wavelet transform in piston error detection. The method first employs wavelet transform to decompose single-channel diffraction images into four components. These components are then processed by a modified CNN (ResNet-18) to extract multichannel features and predict the corresponding piston error value. Simulation results demonstrate that within a single wavelength range, the proposed method significantly outperforms traditional CNN-based methods in terms of detection accuracy and noise robustness. Specifically, under a noise intensity of 20 dB, the proposed WTCNN achieves a root mean square error (RMSE) of 0.002535 lambda , representing approximately a 7.2-fold improvement in detection accuracy compared to traditional CNN methods (RMSE =0.020703 lambda ). Furthermore, experiments under wavefront aberrations and varying noise levels (15-40 dB) reveal the strong generalization capability of the proposed method: it maintains high detection accuracy even when the noise level deviates from the training environment, whereas traditional methods suffer a significant decline in performance. This study not only achieves substantial improvements in the accuracy and noise resistance of piston error detection but also provides a theoretical foundation and practical insights for addressing wavefront aberrations in experimental environments.

With the widespread use of endmember spectral libraries, sparse regression techniques have become crucial in hyperspectral image unmixing. Recently, considering spatial information in sparse unmixing frameworks has become increasingly important, as it enhances the accuracy of mixed pixel decomposition. However, challenges such as spectral mismatch due to high correlation among endmember spectra and susceptibility to complex noise, like Gaussian and sparse noise, affect unmixing accuracy. To address these issues, this paper introduces the robust spatially regularized sparse unmixing algorithm with spectral library pruning (RSUSLP). The algorithm decomposes the unmixing process into multiple layers and prunes the spectral library at each layer to alleviate spectral mismatch. It models sparse noise within the unmixing framework and utilizes spectral weighting in conjunction with spatial weighting to increase row sparsity and spatial correlation, thus improving robustness. The optimization problem described in the algorithm is solved using the alternating direction method of multipliers (ADMM). As illustrated by experimental results from both simulated and real hyperspectral data, RSUSLP significantly surpasses current sparse unmixing methods by reducing spectral library interference and effectively handling noise, thereby enhancing the accuracy and performance of mixed pixel decomposition.

Sparse unmixing has emerged as a powerful technique for addressing the presence of mixed pixels in hyperspectral images. A commonly employed approach involves integrating spectral information and spatial features within a sparse unmixing framework, with the ultimate goal of improving interpretability for remote sensing applications. In practical situations, hyperspectral images frequently encounter various forms of noise contamination, posing challenges to the differentiation of target signals from mixed noise and complicating the interpretation of mixed pixels. To tackle this issue, we introduce a new spatial filtering-based sparse unmixing (SFSU) algorithm. The proposed SFSU model combines filter-guided dual spatial weighting factors and a regularization term to minimize abundance errors within the unmixing framework. By means of this integration, the model facilitates the analysis of spatial information in the image from multiple perspectives, with the objective of mitigating the impact of different noise types on the unmixing outcomes. In the first part of the SFSU model, the original spatial information of the hyperspectral image is preserved by utilizing a precomputed filter-guided spatial weighting factor, thereby augmenting the sparsity of the abundance estimation process. Subsequently, a local neighborhood spatial weighting factor is employed to achieve segmented smoothing in the abundance maps. In the regularization term for reducing abundance errors, multiple filters are introduced to improve the algorithm's robustness to various types of noise by using the difference between the filtered and denoised abundance map and the expected abundance map as a constraint. Extensive experiments conducted on both simulated and real hyperspectral datasets demonstrate the exceptional performance of the SFSU algorithm in effectively addressing complex mixed noise pollution. Moreover, the proposed spatial filtering strategy has been proven to significantly improve the accuracy of unmixing.

The human eye wavefront aberrator based on the Shack-Hartmann wavefront sensor (SHWFS) has become a common device for detecting eye aberrations in modern ophthalmology clinics. In order to eliminate the problem of spot and sub-aperture matching in traditional methods, we use deep learning method to directly map Hartmann spot pattern and corresponding Zernike coefficient, so as to expand the dynamic range of measurement. The lightweight network realizes to fully extract high dimensional feature information and achieves high precision measurement of diopter and astigmatism. The experimental results show that the proportion of the network falling into the tolerant error range (+/- 0.25D) in diopter and astigmatism measurement reaches 94.2% and 100%. This method can measure the low order aberrations of human eyes effectively without changing the SHWFS setting, and at the same time ensure the accuracy and dynamic range, which has been verified by the real machine.

The measurement of the double-pass (DP) point spread function (PSF) provides an objective, non-invasive method for estimating intraocular scatter in the human eye. In this paper, we propose a compact double-pass objective intraocular scatter measurement system that eliminates the influence of aberrations. The system includes a far-field DP PSF detection channel and a Shack-Hartmann wavefront aberration detection channel, which are used to obtain the far-field DP PSF image and 7 orders Zernike aberration coefficients, respectively. The far-field DP PSF image is used to calculate the initial objective scatter index of the human eye. The aberration coefficients are used to reconstruct the DP PSF image caused by aberrations and calculate the influence coefficient of aberrations on intraocular scatter. By subtracting this influence coefficient from the initial objective scatter index (OSI0), the effect of aberrations on scatter measurement can be eliminated, resulting in an accurate objective scatter coefficient. Experimental verification showed that when the exit pupil aperture of this system was set to 4 mm and 6 mm, the measurement accuracy increased by at least 11.9% and 28.9%, respectively, compared to before eliminating the influence of aberrations. While improving the measurement accuracy, the system also keeps the device size and manufacturing costs at a low level, making it more suitable for clinical applications.

A phasing method for the optical multiple-aperture system can be realized using the overlapping pupils when the light beam is focused onto a pyramid, like a pyramid wavefront sensor. The physical idea of this method is to combine beams from different sub-apertures of the multiple-aperture system to form interference based on focal plane filter of the pyramid. When there is a co-phasing piston error between the sub-apertures, the shape of the interference pattern will change significantly, so that the piston error information can be obtained through this change. Theory analysis and simulation results show that this detection method has good accuracy and linearity. The laboratory experiment based on this method was carried out and the results show that the measurement accuracy of the method is within 10nm. Closed-loop correction of large piston error is also realized by using dualwavelength technique to resolve the phase ambiguity and the dynamic range can reach several micrometers.

The resolution of a telescope is closely related to its aperture size; however, the aperture of a single primary mirror telescope cannot be indefinitely enlarged due to design and manufacturing constraints. Segmented mirror technology can achieve the same resolution as a single primary mirror of equivalent aperture, provided that the segments are co-phased correctly. This paper proposes a method for high-precision detection of piston errors in segmented mirror telescope systems, based on far-field information and transfer learning. By training a ResNet-18 network model, this method can predict piston errors with high precision within 10 ms of a single-frame far-field diffraction image. Simulation results demonstrate that the method is robust to tip-tilt errors, wavefront aberrations, and noise. This approach is simple, fast, highly accurate in detection, and resistant to noise, providing a new solution for piston error detection in segmented mirror systems.

Objective To meet the imaging requirements of high resolution and large field-of-view, infrared target detection optical systems should utilize complex optical lens groups, which results in large volume, weight, and total system length. As a result, they are not suitable for deployment on optical payload platforms with limited space, such as airborne and spaceborne platforms. The infrared fiber image bundle is soft and easy to bend, and adding this kind of bundle to the traditional optical system can flexibly change the optical path and shorten the overall length of the system. However, infrared fiber image bundle optical systems have the nature of spatially double discrete sampling effect, and their imaging characteristics are different from those of traditional optical systems. This makes the traditional target detection signal- tonoise ratio (SNR) formula applicable to linear space- invariant systems and no longer applicable to infrared fiber image bundle optical systems. To this end, we present an innovative method for quantitatively analyzing the target detection capability of infrared fiber image bundle optical systems. The proposed method adopts statistical analysis to complete the derivation of the target SNR and the SNR attenuation coefficient formulas, featuring clarity, simplicity, and easy calculation. We hope that this method will contribute to the optical design, device selection, and system detection capability analysis of infrared fiber image bundle optical systems. Methods The target detection scenario of infrared fiber bundle optical systems is set to detect distant point targets under a uniform background in the sky. Thus, the signal and noise components are appropriately corrected respectively to obtain the target detection SNR formula of these systems. First, the overall transmittance of infrared fiber bundle optical systems is obtained by combining the product of the fiber transmittance and the fiber bundle filling factor, the transmittance of the front telescopic system, and the transmittance of the rear coupling system. Then, the proportion of the cross area between the core area of the fiber bundle and the photosensitive area of the detector pixel is defined as the system filling factor, which is employed to characterize the spatially double discrete sampling effect. By utilizing the overall transmittance and filling factor of optical systems, the noise equivalent temperature difference can be statistically derived, and then the noise equivalent power can be obtained, which represents the noise component of infrared fiber bundle optical systems. For the signal component correction, the introduction of the pulse visibility factor is to describe the energy concentration of infrared fiber bundle optical systems on the point target image. Based on the correction formulas for the above-mentioned noise and signal components, a target SNR formula for infrared fiber bundle optical systems is derived, which includes target radiation characteristics, background radiation characteristics, optical system parameters, and detector parameters. To simplify the analysis of system detection capability, we define the SNR attenuation coefficient as the proportion of SNR decrease in infrared fiber bundle optical systems compared to traditional optical systems. Finally, combined with the derived SNR attenuation coefficient and designed structural parameters of the infrared fiber bundle optical system, the system filling factor and pulse visibility factor are calculated, and the relationship between the fiber transmittance and the SNR attenuation coefficient is given. Finally, this can quantitatively evaluate the difference in the detection ability of the infrared fiber bundle optical system. Results and Discussions In the condition of vertical coupling alignment assembly (or matching a certain column of vertical fiber bundle images with a square pixel line array), we combine the optical design results of the point spread function (PSF) of the front telescopic system and the rear coupling system ( Fig. 8), and the fiber bundle characteristic function distribution (Fig. 9). Meanwhile, the average pulse visibility factor of the infrared fiber bundle optical system with the fiber bundle resolution of 25x256 toward the point target is calculated to be 0.1335. Due to the coupling mismatch between the infrared fiber bundle and the square pixel array, the fiber coupling area varies for different pixels ( Fig. 13), with the calculated average filling factor of 0.4201. Based on the calculated pulse visibility factor and system filling factor of the infrared fiber bundle optical system and the traditional optical system, the relationship between the detection SNR attenuation coefficient ASNR and the fiber transmittance tfiber is given (Fig. 14), and the following conclusions can be drawn. Under tau(fiber) > 0.9, A(SNR) < 0.5; under tau(fiber) < 0.3, A(SNR) > 0.7; under tau(fiber) < 0.03, A(SNR) > 0.9. Therefore, it is advisable to adopt fiber bundle devices with high transmittance to improve the detection capability of fiber bundle systems. Conclusions We propose an innovative quantitative analysis method for the ability of infrared fiber bundle optical systems to detect distant targets, thereby solving the problem that traditional target SNR formulas are not applicable to such optical systems with spatially double discrete sampling effect. Based on the imaging theory of infrared fiber bundle optical systems, the target SNR formula is derived by appropriately modifying the signal expression and noise expression. Additionally, the SNR attenuation coefficient expression of the system compared to traditional optical systems is provided, which can effectively characterize the detection ability of the system. Based on the designed infrared fiber bundle optical system, key performance parameters such as the pulse visibility factor and system filling factor are simulated and calculated. The relationship between the fiber transmittance and SNR attenuation coefficient is further analyzed, with the detection ability differences between the two types of optical systems quantitatively compared. The simulation results demonstrate the influence of the spatially double discrete sampling effect on the infrared fiber bundle optical system and clarify that the SNR attenuation coefficient is related to a fixed coefficient of 0.5459 and fiber transmittance. Thus, it is indicated that the detection capability of the system can be improved by selecting fiber bundle devices with high transmittance. Finally, we can provide a theoretical basis for determining the detection ability boundary of infrared fiber bundle optical systems.

We have established a novel spatial heterodyne spectroscopy (SHS) signal-to-noise ratio (SNR) model, relating the spectral SNR to the spectral band and resolution, which helps guide the instrument's design and optimization and assess the spectral quality. The experimental and simulation results show that the spectral resolution affects the SNR differently under different noise types and spectral characteristics of the measured targets, which fully validates the novel SHS SNR model. Despite the disadvantage of multiplexing in SHS, we determine the SNR advantage of SHS over the grating spectroscopy (GS) through rigorous theoretical derivations. In 94.34 % of the polychromatic light detections, the average SNR of SHS is 2-31 times higher than that of GS. In emission spectra detections, the SNR gain of SHS relative to GS is up to 1-2 orders of magnitude. In additive noise domination, the SNR gain reaches 1-3 orders of magnitude.

There are more than 40 years history of adaptive optics (AO) in Institute of Optics and Electronics (IOE), Chinese Academy of Sciences since 1980. The research concern all the aspects including the theories study, devices manufacture, and system development. The recent advances on astronomical AO are reported in this presentation. The recent AO systems developments for 4-meter night-time optical telescope, 1.8-meter solar telescope CLST and the 1-m New Vacuum Solar Telescope at Fuxian Lake Solar Observatory are presented respectively. The Deformable Secondary Mirror advancement is also introduced.

Spatial heterodyne one-dimensional imaging spectrometer (SHIS) can simultaneously acquire hyperspectral information from different fields of view (FOVs). However, the dynamic range of SHIS is limited by the detector's performance. We propose a high dynamic range spatial heterodyne one-dimensional imaging spectroscopy (HD-SHIS) based on a digital micromirror device (DMD), which can control the exposure time of each FOV signal by adjusting the flip time of micromirrors on an M-bit DMD, realizing the simultaneous detection of strong and weak signals in FOVs with a theoretical improvement of the dynamic range by dB. Meanwhile, HD-SHIS utilizes a DMD to realize the Hadamard modulation of interference data in the spectral dimension, which can be used with the linear array detector to complete the detection of the imaging spectrum. We have built an HD-SHIS principle prototype and carried out dynamic range experiments. The experimental results show that HD-SHIS can achieve 48 dB dynamic range improvement by utilizing an 8-bit display width DMD.

Improving the spectral signal-to-noise ratio is crucial for obtaining high-quality Raman spectra of target substances. To achieve a higher instrument signal-to-noise ratio in the development of spatial heterodyne Raman spectrometer, the effect of spatially compressed illumination on the spectral signal-to-noise ratio was theoretically analyzed. A spatial heterodyne Raman spectroscopic experimental setup with compressed illumination was constructed. The experimental results with ethanol validate the significant enhancement of the signal-to-noise ratio achieved through compressed illumination. The ethanol spectra exhibited signal-to-noise ratio enhancements ranging from 1.38 to 7.98 times under various excitation powers and integration times. These findings provide valuable guidance for further improving the performance of spatial heterodyne Raman spectrometers.

In the adaptive optics system of large-aperture ground-based telescopes, the wavefront sensor plays a crucial role. Pyramid wavefront sensors are increasingly favored by an expanding number of world-class telescopes. However, traditional wavefront reconstruction algorithms with pyramid wavefront sensors have limited ability to fit nonlinearity, resulting in restricted improvement in reconstruction accuracy. The kernel of deep learning lies in the ability of artificial neural networks to approximate nonlinear functions with arbitrary precision, which is well-suited for solving the nonlinear wavefront reconstruction problem of pyramid wavefront sensors and achieving more accurate wavefront sensing. This paper introduces the application of deep learning in pyramid wavefront sensors and Shack-Hartmann wavefront sensors, conducts a comparative analysis between them, and discusses potential future research directions.

Aiming for long-distance, high-resolution, passive imaging, we use fiber coupling to replace the spatial coupling of the segmented planar imaging detector for electro-optical reconnaissance imaging system and propose a novel incoherent interference imaging method, which avoids the problem that the resolution of the imaging system is limited by the size of the silicon wafer. The method uses a phase retrieval algorithm and does not need to measure the phase in the system, effectively avoiding the intrinsic jitter problem of fiber optics. The feasibility of the method in achieving target image reconstruction was verified through simulation; furthermore, an indoor two-dimensional discrete sampling imaging experiment for a simple four-rod target was conducted, and an outdoor one-dimensional feature identification experiment for a two-rod target was also performed. The results showed that the resolution exceeds the diffraction limit and the reconstruction error of target size is less than 3.5%.

Future segmented space telescope needs to work at larger aperture and shorter wavelength, and the co-phasing state will rapidly deteriorate due to environmental interference. To avoid co-phasing drift during the observation period, it is necessary to use edge sensors for real-time correction of phasing error. The unshielded edge sensors installed on the primary mirrors are exposed to the outermost place of the satellite. During working periods of more than ten years, the impact of irradiation effects on sensors is particularly obvious. We have developed a radiation resistant capacitive edge sensor with foundry-based components. After a total ionizing dose (TID) test of 1.6 x 10(9) rad (Si), analyses were conducted on the appearance, roughness, surface composition, detection range, and accuracy. The results show that the relative error of the irradiated sensor within the displacement range of 6 nm- 20 mu m is between 0.85% and 7.65%, indicating that the irradiated sensor still meets the requirements for detection range and accuracy.

The Shack–Hartmann wavefront sensor (SHWFS) is widely utilized for ocular aberration measurement. However, large ocular aberrations caused by individual differences can easily make the spot move out of the range of the corresponding sub-aperture in SHWFS, rendering the traditional centroiding method ineffective. This study applied a novel convolutional neural network (CNN) model to wavefront sensing for large dynamic ocular aberration measurement. The simulation results demonstrate that, compared to the modal method, the dynamic range of our method for main low-order aberrations in ocular system is increased by 1.86 to 43.88 times in variety. Meanwhile, the proposed method also has the best measurement accuracy, and the statistical root mean square (RMS) of the residual wavefronts is 0.0082 ± 0.0185 λ (mean ± standard deviation). The proposed method generally has a higher accuracy while having a similar or even better dynamic range as compared to traditional large-dynamic schemes. On the other hand, compared with recently developed deep learning methods, the proposed method has a much larger dynamic range and better measurement accuracy.