RESUMO
Since light propagation in water bodies is subject to absorption and scattering effects, underwater images using only conventional intensity cameras will suffer from low brightness, blurred images, and loss of details. In this paper, a deep fusion network is applied to underwater polarization images; that is, the underwater polarization images are fused with intensity images using the deep learning method. To construct a training dataset, we establish an experimental setup to obtain underwater polarization images and perform appropriate transformations to expand the dataset. Next, an end-to-end learning framework based on unsupervised learning and guided by an attention mechanism is constructed for fusing polarization and light intensity images. The loss function and weight parameters are elaborated. The produced dataset is used to train the network under different loss weight parameters, and the fused images are evaluated based on different image evaluation metrics. The results show that the fused underwater images are more detailed. Compared with light intensity images, the information entropy and standard deviation of the proposed method increase by 24.48% and 139%. The image processing results are better than other fusion-based methods. In addition, the improved U-net network structure is used to extract features for image segmentation. The results show that the target segmentation based on the proposed method is feasible under turbid water. The proposed method does not require manual adjustment of weight parameters, has faster operation speed, and has strong robustness and self-adaptability, which is important for research in vision fields, such as ocean detection and underwater target recognition.
Assuntos
Algoritmos , Aprendizado de Máquina não Supervisionado , Benchmarking , Entropia , ÁguaRESUMO
This article studies the measurement error model and calibration method of the bio-inspired polarization imaging orientation sensor (BPIOS), which has important engineering significance for promoting bio-inspired polarization navigation. Firstly, we systematically analyzed the measurement errors in the imaging process of polarized skylight and accurately established an error model of BPIOS based on Stokes vector. Secondly, using the simulated Rayleigh skylight as the incident surface light source, the influence of multi-source factors on the measurement accuracy of BPIOS is quantitatively given for the first time. These simulation results can guide the later calibration of BPIOS. We then proposed a calibration method of BPIOS based on geometric parameters and the Mueller matrix of the optical system and conducted an indoor calibration experiment. Experimental results show that the measurement accuracy of the calibrated BPIOS can reach 0.136°. Finally, the outdoor performance of BPIOS is studied. Outdoor dynamic performance test and field compensation were performed. Outdoor results show that the heading accuracy of BPIOS is 0.667°.
RESUMO
With the vigorous development of vision techniques, simultaneous localization and mapping (SLAM) has shown the capability of navigating autonomous robots in global-navigation-satellite-system-denied environments. However, the long-term robust navigation of lightweight autonomous robots in outdoor environments with complex interferences, such as illumination change, dynamic objects, and electromagnetic interference, is still a great challenge. In this paper, a polarization sensor-aided SLAM (POL-SLAM) that can provide absolute heading constraints for pure SLAM is proposed. POL-SLAM is a lightweight, tightly coupled system consisting of a polarization sensor and binocular camera. By means of an initialization that uses a polarization sensor, an absolute heading angle for the entire map is designed. Additionally, an algorithm to eliminate mismatching points using the matching point vector is proposed. The objective function of bundle adjustment is then deduced according to the re-projection error and polarization sensor. The vehicle test shows that the yaw and trajectory accuracies of POL-SLAM are significantly improved compared to pure SLAM. The yaw and trajectory accuracies are increased by 43.1% and 36.6%, respectively. These results indicate that the proposed POL-SLAM can improve the reliability and robustness of pure SLAM and can be used in lightweight autonomous robots in outdoor environments.
Assuntos
Robótica , Algoritmos , Biomimética , Reprodutibilidade dos Testes , Robótica/métodosRESUMO
Underwater imaging faces challenges due to complex optical properties in water. Our purpose is to explore the application of polarimetric imaging in image recovery under turbid water based on deep learning. A polarization camera is used to capture the polarization images of objects under water as datasets. The method used in our study aims to explore a structure and loss function that is suitable for the model. In terms of the model structure, four pairs of models consisting of polarized version and gray version based on the idea of dense U-Net and information flow were proposed. In the aspect of loss function, the method of combining weighted mean squared error with perceptual loss was proposed and a proper set of loss weights was selected through comparison experiments. Comparing the model outputs, it is found that adding polarized information along with the light intensity information to the model at the very front of the model structure brings about better recovering image. The model structure proposed can be used for image recovery in turbid water or other scattering environments. Since the polarization characteristics are considered, the recovered image has more detailed features than that where only intensity is considered. The results of comparison with other methods show the effectiveness of the proposed method.
RESUMO
High precision and reliability attitude measurement play an important role in autonomous unmanned navigation. Finding inspiration from desert ants, known as highly efficient navigators who can find their way after foraging for hundreds of meters from their home in hostile environments, we propose an attitude measurement method using polarization skylight and gravitational field. Contrary to the previous method, we utilize three-dimensional polarization vectors and any one-dimensional output of the accelerometers to calculate attitudes. In addition, we designed an accelerometer component selection algorithm, which is to select the one-dimensional component with the minimum motion acceleration from the output of the three-dimensional accelerometer. With this method, even if the carriers remain in a maneuvering state, the motion acceleration of the vehicle will have less impact on the accuracy of attitude measurement. To evaluate the performance of our method, the outdoor experiment was carried out to compare our method with existing traditional methods. Comparison results show that our method has higher measurement accuracy than others and is still applicable in the case of carriers maneuvering in practice under a clear sky.
RESUMO
Polarized light forward propagation in scattering environments is important basic research. Polystyrene microspheres in water are common scattering environments that can be helpful to investigate in existing literature research. In this paper, we investigated the polarization state persistence of both linearly and circularly polarized light. We used a single active source with a wavelength of 532 nm to illuminate 1 µm diameter polystyrene spheres immersed in water. To evaluate the polarization state persistence of linearly and circularly polarized light, a parameter change of polarization state was used to replace the Stokes parameters. In the setting environments of different concentrations, circularly polarized light has superior polarization state persistence to that of linearly polarized light.
RESUMO
We fabricated a cost-effective mid-IR micropolarizer on a common Si substrate. To improve the transmittance of Si, we performed a double oxidation on the silicon substrate. The SiO2-Si-SiO2 structure improved the transmittance of Si from 54% to 63%-83%. Then, the mid-IR micropolarizer with multidirectional gratings was fabricated using a soft nanoimprint process followed by the thermal evaporation of Al. Experimental measurements showed a transverse magnetic transmittance in the range of 61%-80% at wavelengths of 4-5 µm, and the extinction ratio was greater than 19 dB.
RESUMO
Consistency has always been an important topic in formation cooperation research. Traditional navigation methods, such as inertial navigation and geomagnetic navigation, have the disadvantages of error accumulation and low stability, thus reducing the consistency of formation. We propose to use the skylight polarization pattern to provide heading angle reference for formation cooperation of multi-agents. The experimental results show that the polarization navigation has good stability and no error accumulation. First, we analyzed the consistency of using the skylight polarization pattern to provide a heading reference for formation experiments. Then, based on the bionic polarization navigation sensor, we measured the difference of the skylight polarization azimuth of different observers at twilight. Further, a mobile robot platform was built with its heading angle provided by a polarization navigation sensor. Finally, we present an overview of a 3-robots platform formation experiment at twilight.
RESUMO
A bio-inspired polarization sensor with lenses for navigation was evaluated in this study. Two new calibration methods are introduced, referred to as "central-symmetry calibration" (with an integrating sphere) and "noncontinuous calibration". A comparison between the indoor calibration results obtained from different calibration methods shows that the two proposed calibration methods are more effective. The central-symmetry calibration method optimized the nonconstant calibration voltage deviations, caused by the off-axis feature of the integrating sphere, to be constant values which can be calibrated easily. The section algorithm proposed previously showed no experimental advantages until the central-symmetry calibration method was proposed. The outdoor experimental results indicated that the indoor calibration parameters did not perform very well in practice outdoor conditions. To establish the reason, four types of calibration parameters were analyzed using the replacement method. It can be concluded that three types can be easily calibrated or affect the sensor accuracy slightly. However, before the sensor is used outdoors every time, the last type must be replaced with the corresponding outdoor parameter, and the calculation needs a precise rotary table. This parameter, which is mainly affected by the spectrum of incident light, is the main factor determining the sensor accuracy. After calibration, the sensor reaches an indoor accuracy of ±0.009° and a static outdoor accuracy of ±0.05° under clear sky conditions. The dynamic outdoor experiment shows a ±0.5° heading deviation between the polarization sensor and the inertial navigation system with a ±0.06° angular accuracy.
RESUMO
Detection range is an important factor affecting the transmission characteristics of polarized light through fog. We first selected certain spectral bands from visible to IR wavelengths that exhibit lower path loss. For both radiation fog and advection fog, these optimized wavelength ranges include 0.4-1.1 µm, 1.48-1.56 µm, 1.63-1.86 µm, 2.03-2.18 µm, and 2.39-2.45 µm, and radiation fog in particular contains 3.5-4.3 µm. The long-wave IR wavelengths were excluded due to higher absorption losses. We further investigated the transmission performance of circular and linear polarization in variable foggy environments, exploring the impact of the detection range in particular. Using polarization-tracking Monte Carlo simulations for varying particle size, wavelength, refractive index, and detection range, we show that circular polarization outperforms linear polarization when transmitting in both radiation and advection fog. For radiation fog, circular polarization persists longer than linear polarization for 5 µm and 9 µm particles over the entire optimized wavelength range from the visible to mid-wave IR (MWIR). However, linear polarization outperforms circular polarization for 1 µm particles over the entire MWIR and a part of the short-wave IR (SWIR). For advection fog, circular polarization persists longer than linear polarization for all three particle sizes (10, 20, and 40 µm) over the entire optimized wavelength range from the visible to SWIR. We show that circular polarization retains a higher degree of polarization and has better enhancement in some detection ranges.
RESUMO
We fabricated a 4-in large-area flexible infrared nanowire grid polarizer using a nanoimprint and metal thermal evaporation process. To protect the Si master template, as well as to prolong the service life of it, we first fabricated a nickel template as an alternative by an electroforming process. Then, the nanowire grid structure was transferred from this template to IPS substrate by a thermal nanoimprint process. Finally, Al was deposited on the IPS nanowire structure by vertical thermal evaporation technology. The results of the infrared optical test reveal that the TM transmittance of the polarizer is greater than 60% in the 4-5.71 µm and 5.73-6.7 µm wavelength ranges, and, especially, it is greater than 70% in the wavelength ranges of 4.70-5.69 µm and 5.75-6.59 µm. The extinction ratio is more than 20 dB in the wavelength range of 3.6-6.7 µm, proving that the polarizer has good polarization characteristics. The flexible infrared nanowire grid polarizer has potential applications in the fields of curved surface monitoring equipment and polarized imaging equipment.
RESUMO
Information obtained from a polarization pattern in the sky provides many animals like insects and birds with vital long-distance navigation cues. The solar vector can be derived from the polarization pattern using the single-scattering Rayleigh model. In this paper, an orthogonal vector algorithm, which utilizes the redundancy of the single-scattering Rayleigh model, is proposed. We use the intersection angles between the polarization vectors as the main criteria in our algorithm. The assumption that all polarization vectors can be considered coplanar is used to simplify the three-dimensional (3D) problem with respect to the polarization vectors in our simulation. The surface-normal vector of the plane, which is determined by the polarization vectors after translation, represents the solar vector. Unfortunately, the two-directionality of the polarization vectors makes the resulting solar vector ambiguous. One important result of this study is, however, that this apparent disadvantage has no effect on the complexity of the algorithm. Furthermore, two other universal least-squares algorithms were investigated and compared. A device was then constructed, which consists of five polarized-light sensors as well as a 3D attitude sensor. Both the simulation and experimental data indicate that the orthogonal vector algorithms, if used with a suitable threshold, perform equally well or better than the other two algorithms. Our experimental data reveal that if the intersection angles between the polarization vectors are close to 90°, the solar-vector angle deviations are small. The data also support the assumption of coplanarity. During the 51 min experiment, the mean of the measured solar-vector angle deviations was about 0.242°, as predicted by our theoretical model.
RESUMO
Using a reversal nanoimprint and metal evaporation process, we fabricated a micropolarizer array for the 2.5-7 µm wavelength region. The micropolarizer array has a unique unit, which is composed of 2×3 arrays on an intrinsic silicon substrate. Each array consists of a 200 nm period bilayer Al grating in a 1.3 mm×1.3 mm aperture. The transmittance of transverse magnetic polarization of each array is greater than 65% in the 2.5-7 µm wavelength range, and the extinction ratio is over 35 dB in the 3-4 µm and 6-7 µm wavelength range. This fabricated micropolarizer array has lower costs and better compatibility with microfabrication processes.
RESUMO
Polarized skylight navigation has excellent navigation performance with no error accumulation over time and low susceptibility to interference. The skylight polarization distribution contains rich directional information, such as the solar meridian, the neutral point, and the polarization angle, which plays a key role in the polarization navigation. But up to now the polarizations of both sunlit and moonlit skies have been investigated mainly over the land. In this work, the polarization distribution patterns of the skylight over the East China Sea and the Yellow Sea were studied. The polarization patterns were captured continuously during daytime and nighttime by using a full-sky imaging polarimetry system and then compared with the simulation results using the libRadtran radiative transfer software package. The result shows that the skylight polarization distribution over the sea has almost the same pattern as that on the land. The accuracy of the angle of polarization and the degree of polarization dropped significantly under the cloudy sky. It was found that when the ship sailed on the sea, the direction of the real meridian was close to the solar azimuth during the daytime and close to the lunar azimuth during the nighttime. It was also found that the nautical twilight polarization distribution was affected by both the solar polarization and the lunar polarization, but the solar polarization was dominant. The experiments show that the skylight polarization distribution pattern over the sea can still be applied in the field of polarization navigation. Thus, it is feasible for ships and unmanned aerial vehicles to use the polarized skylight to navigate and orient on the sea.
RESUMO
This paper aims to develop a novel attitude determination system aided by polarization sensor. An improved heading angle function is derived using the perpendicular relationship between directions of E-vector of linearly polarized light and solar vector in the atmospheric polarization distribution model. The Extended Kalman filter (EKF) with quaternion differential equation as a dynamic model is applied to fuse the data from sensors. The covariance functions of filter process and measurement noises are deduced in detail. The indoor and outdoor tests are conducted to verify the validity and feasibility of proposed attitude determination system. The test results showed that polarization sensor is not affected by magnetic field, thus the proposed system can work properly in environments containing the magnetic interference. The results also showed that proposed system has higher measurement accuracy than common attitude determination system and can provide precise parameters for Unmanned Aerial Vehicle (UAV) flight control. The main contribution of this paper is implementation of the EKF for incorporating the self-developed polarization sensor into the conventional attitude determination system. The real-world experiment with the quad-rotor proved that proposed system can work in a magnetic interference environment and provide sufficient accuracy in attitude determination for autonomous navigation of vehicle.
RESUMO
Mueller matrix imaging polarimetry (MMIP) is a promising technique for investigating structural abnormalities in pathological diagnosis. The characterization stability of polarization signatures, described by Mueller matrix parameters (MMPs), correlates with the mechanical state of the biological medium. In this study, we developed an MMIP system capable of applying quantitative forces to samples and measuring the resulting polarization signatures. Mechanical stretching experiments were conducted on a mimicking phantom and a tissue sample at different force scales. We analyzed the textural features and data distribution of MMP images and evaluated the force effect on the characterization of MMPs using the structural similarity index. The results demonstrate that changes in the mechanical microenvironment (CMM) can cause textural fluctuations in MMP images, interfering with the stability of polarization signatures. Specifically, parameters of anisotropic orientation, retardance, and optical rotation are the most sensitive to CMM, inducing a dramatic change in the overall image texture, while other parameters (e.g., polarization, diattenuation, and depolarization) exhibit locality in their response to CMM. For some MMPs, CMM can enhance regional textural contrasts. This study elucidates the mechanical stability of polarization signatures in biological tissue characterization and provides a valuable reference for further research toward minimizing CMM influence.
RESUMO
Nitrogen dioxide (NO2) is a significant environmental and human health hazard. Current NO2 sensors often lack sensitivity and selectivity under ambient conditions. This study investigates ammonia pyrolysis modification of monolayer Ti3C2Tx MXene to enhance NO2 detection at room temperature. Nitrogen-doped Ti3C2Tx demonstrates a substantial improvement in sensitivity, with a response of 8.87% to 50 ppm of NO2 compared to 0.65% for the original sensor, representing a 13.8-fold increase. The nitrogen-doped sensor also exhibits superior selectivity and linearity for NO2 under ambient conditions. Theoretical analysis shows that nitrogen incorporation promotes enhanced interaction between Ti3C2Tx and its surface oxygen-containing functional groups through electronic hybridization, resulting in improved adsorption energy (1.80 |eV|) and electron transfer efficiency (0.67 |e|) for NO2, thereby enhancing its gas-sensing performance. This study highlights the potential of ammonia pyrolysis-treated Ti3C2Tx MXene for advancing NO2 sensor technologies with heightened performance at room temperature.
RESUMO
Chiral metamaterials are extensively applied in the fields of photoelectric detection, biomedical diagnostics and micro-nano polarization imaging. Currently, single-layer chiral metamaterials are unfortunately limited by several issues, such as a weaker circular polarization extinction ratio and circular polarization transmittance difference. To tackle these issues, a single-layer transmissive chiral plasma metasurface (SCPMs) suitable for visible wavelength is proposed in this paper. Its basic unit is composed of double orthogonal rectangular slots and a spatial π/4 inclined arrangement of the rectangular slot to constitute a chiral structure. Each rectangular slot structure has characteristics that enable the SCPMs to easily achieve a high circular polarization extinction ratio and strong circular polarization transmittance difference. Both the circular polarization extinction ratio and circular polarization transmittance difference of the SCPMs reach over 1000 and 0.28 at a wavelength of 532 nm, respectively. In addition, the SCPMs is fabricated via the thermally evaporated deposition technique and focused ion beam system. This compact structure coupled with a simple process and excellent properties enhances its applicability for the control and detection of polarization, especially during integration with linear polarizers, to achieve the fabrication of a division-of-focal-plane full-Stokes polarimeter.
RESUMO
A new technique of polarization doping was adopted to improve NO2 gas sensing properties of the polypyrrole (PPy) sensor. PPy nanosheets polarization doped with sodium dodecyl benzenesulfonate (SDBS) were synthesized by low-temperature polymerization. The semiagglomerated PPy nanosheets were well-dispersed and a large specific surface areas due to the introduction of dodecyl benzenesulfonate (DBS). The DBS doped PPy sensor shows excellent NO2 sensing performance. Polarization of sulfonate ions to PPy enhanced the adsorption ability of NO2 with the synergistic effect of NO2. The adsorption energy (-0.676 eV) and electron transfer (0.521 |e|) of PPy to NO2 increased greatly after doping. An unoccupied electron state is observed in the valence band electron structure of PPy/DBS after the adsorption of NO2 by calculations of Density Functional Theory (DFT). The intermolecular synergy between NO2 and PPy/DBS causes a strong polarization, resulting in a high polarization potential, which enhances the NO2 sensing performance of PPy sensor. It is of great significance to develop NO2 detection device based on PPy that works at room temperature.
RESUMO
Mueller matrix imaging polarimetry (MMIP) is a promising technique for the characterization of biological tissues, including the classification of microstructures in pathological diagnosis. To expand the parameter space of Mueller matrix parameters, we propose new vector parameters (VPs) according to the Mueller matrix polar decomposition method. We measure invasive bladder cancer (IBC) with extensive necrosis and high-grade ductal carcinoma in situ (DCIS) with MMIP, and the regions of cancer cells and fibrotic stroma are classified with the VPs. Then the proposed and existing VPs are mapped on the Poincaré sphere with 3D visualization, and an indicator of spatial feature is defined based on the minimum enclosing sphere to evaluate the classification capability of the VPs. For both IBC and DCIS, the results show that the proposed VPs exhibit evident contrast between the regions of cancer cells and fibrotic stroma. This study broadens the fundamental Mueller matrix parameters and helps to improve the characterization ability of the MMIP technique.