This PDF file contains the front matter associated with SPIE Proceedings Volume 11428, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

In this paper, we propose a novel dual-function infrared liquid-crystal device (DF-ILCD), which can simultaneously perform both tunable focusing and filtering functions through applying alternating current (AC) voltage signals. The key functional micro-structure of the DF-ILCD includes: two paralleled 1-mm-thick ZnSe substrates with 20-nm-thick aluminum (Al) films over their inner surfaces and a periodic repetition microholes over both substrates. The Al films act as both high-reflection films and conductive films. The conventional UV-photolithography and wet-etching process are used to fabricate an arrayed micro-hole with a diameter of 120μm and a period of 336μm over each substrate. The micro-cavity formed between ZnSe substrates has a typical depth of ~12 μm, which is fully filled by a nematic liquid-crystal (LC) materials. Experiments demonstrate that the proposed device exhibit both filtering function based on Fabry–Perot (FP) effect and electrically controlled tunable focusing function generated by the micro-hole electrodes, which is very promising for realizing both the light-field imaging and spectral imaging in the infrared wavelength range.

As an effective method for collecting light field information and further extending the depth of field, a new imaging technology based on key electrically controlled liquid-crystal microlens array (EC-LCMLA), has been proposed. Compared with common lenses with defined surface profile, the liquid-crystal microlenses can be used to regulate the focal length only through applying different signal voltages to achieve focus tuning or even swing on the observation plane. Generally, the traditional autofocus operations are no longer suitable to EC-LCMLA because the controlling orders for LC structures should be generated through image process. So, an autofocus method, which is used to dynamically adjust the focal length of each imaging unit in the EC-MLA, is proposed for controlled LCMLA in this paper. The method is used to extract the light field information from low-quality image, so as to obtain the key focusing distance of the plane observed by each imaging unit, and then calculate the focal length of the EC-LCMLA without additional sensors. The signal voltage of each liquid-crystal microlens can be adjusted by the driving control unit, which implements an automatic focusing of the LCMLA. The active autofocus therefore is achieved and then all the imaging units in an optimal working state. Based on theoretical analysis and the focusing algorithm constructed by us, the imaging experiments are carried out so as to show a higher performance and then image quality and focusing efficiency of LCMLA. The novel autofocus method highlights a construction of a new kind of plenoptic camera with stronger performances.

A planar-array multiple-input-multiple-output (MIMO) radar system possess the ability to gain a three-dimensional (3-D) image in single snapshot due to the wide-band signal and two-dimensional (2-D) virtual aperture. And the conventional inverse synthetic aperture radar (ISAR) obtains the cross-range resolution thanks to the relative rotational movement during the observation. Naturally, the planar-array MIMO radar 3-D images in multiple snapshots also include slow-time domain Doppler information. In order to take advantage of the Doppler shift along the slow-time domain for a better 3-D imaging result, we investigate the method of MIMO radar 3-D imaging via jointly utilizing the time-space observation. By coherent processing along the velocity direction, inverse aperture caused by target movement is incorporated into the 3-D image focusing and therefore the resolution can be increased. Simulation results validate the effectiveness of the proposed method. Comparing to ISAR, the longtime observation as well as the complicated motion compensation in the proposed 3-D imaging method is not necessary. Besides, comparing to the 3-D image in single snapshot, the proposed method can improve the resolution along the target trajectory efficiently.

Due to the large dynamic range and low contrast, inverse synthetic aperture radar (ISAR) image is not appropriate for human observation. In order to output and display the target imaging results, a procedure which compresses the dynamic range of the raw images into a lower range is necessary. In this paper, by analyzing the histogram of original ISAR images, the characteristics of ISAR images are investigated. Given the sparse amplitude distribution of original ISAR image and the shortcomings existing in the sparse linear histogram, this paper proposes an ISAR image detail enhancement algorithm (ISARIDE) based on histogram equalization and dynamic range compression. The advantage of the proposed method is that it can retain the target structure information, and improve the visual effect of human eye to target details as soon as possible. The proposed algorithm was tested on the simulated data and real data. The selected target is a flying Boeing 737-800.The results show the validity of the algorithm.

In order to improve the inverse synthetic aperture radar (ISAR) imaging quality of precession space target, an algorithm based on phase matching processing (PMP) of complex range profile envelope (CRPE) is proposed in this paper. By phase matching processing, only the echo components located at the scattering points corresponding to the main body of the warhead can be coherently accumulated. The echoes of other scattering centers without any coherence will cancel each other when they are transformed. Hence, the focusing of scatters centers of the warhead main body is improved and the interference caused by non-spin symmetric components is well suppressed. Simulation results confirmed the effectiveness of the method.

So far, how to acquire an effective method of integration detection covering a relatively wide wavelength range has become a hot topic in the field of high performance radiation detection. In this work, the microstructure patterned Schottky-typed optical antenna is designed and then fabricated on Gallium Arsenide substrate and further used to sense near-infrared lights and terahertz signals, respectively. The wide frequency terahertz waves generated by InAs crystal are measured through patterned optical antenna device, and then the characteristics of transmitted waves are analyzed, it should be noted that the time delay characteristics of transmitted terahertz signals between 5ps and 8ps are different from microstructure patterned optical antenna. Under the conditions of using near-infrared lasers and also adjusting main parameters such as the exposure time, for example, 0.04ms、0.4ms、0.6ms、0.8ms、1.0ms and 1.5ms, in the experiments, the transmitted image characters acquired using functioned optical antennas with different electrode patterns, are analyzed. In the near-infrared transmission experiments, the transmitted bright light points or spots with relatively large distribution density and high intensity and very small structural size (~1μm), are discovered, which distribute over the top layer of electrode zone without metal structures of optical antenna device. The developed detection architecture based on Schottky-typed functioned optical antennas to sense infrared light and terahertz radiation, is expected to integrated sense electromagnetic signals in wide spectrum regime.

Metallic micro-nano-structure arrays can be used to induce a collective oscillation of free electrons on the surface of metal films, so as to generate relatively strong surface plasmons (SPs) at the metal and medium interface and further localized light field under the excitation of incident lightwaves. As the oscillating light field propagating along the interface, the field strength can be increased reasonably at the functioned metal surface such as the incident light energy being localized in the sub-wavelength region defined by the functioned micro-nano-structures. The common beam diffraction limit formed during lightwave transmission or process can be broken effectively. Through constructing SPs over the special micro-nano-structures, the infrared reflection characteristics can be changed and then the local light field originated from incident infrared radiation also be enhanced significantly so as to efficiently perform infrared detection. Generally, the reflectivity and light field distribution behaviors of the functioned metal surface can be modulated by changing featured parameters of the metallic micro-nano-structural arrays. In this paper, a metal micro-nano-patterned structures with an arrayed tip is established for compressing the incident light field and then reducing the reflectivity of the metal surface and thus sensing incident light energy. A finite integral method for simulating and analyzing the structural characters such as the distance between tips, the tip sharpness, the thickness of the metal film, is utilized to acquire the reflectivity and field enhancement characteristics. The infrared reflection spectrum and the near-field intensity distribution of the metallic micro-nano-structure are compared and analyzed. The results show that the response frequency and excitation intensity of SPs over the nano-tip array, the intensity and distribution region of the strong light field, can be controlled by matching the structural parameters and layout. The optimization of the metallic micro-nanostructure arrays is conducted so as to lay a solid foundation for further development of the similar technologies.

Generally, the vortex lightwaves are beams that propagates helically along an axis. In an optical vortex, light is twisted. Due to their twisting property, the light at the position of the axis cancels each other. If the vortex is projected onto a plane, its image looks like a halo, with a dark area without light in the middle. Vortex beams are a new type of laser beams, which have been widely used in optical micromanipulation, laser optics, bio-optics, optical information transmission, particle waveguide, atomic optics, and molecular optics. Compared to ordinary Gaussian beams, vortex beams have many advantages, such as: zero intensity at the center of the beam, cylindrical distribution of longitudinal light intensity, dark spot size on the order of micrometers, no heating effect, orbital angle momentum and so on. This makes the vortex beams have very high application value in many fields, such as optical capture, optical rotation, nonlinear optics, and optical information processing. The scientific value of the vortex beams is increasingly prominent. In this paper, a novel structure similar to a spiral phase plate(SPP) to generate a vortex beam is designed, and its intensity distribution is analyzed using simulation software, which can be seen as a combination of a lens and a spiral phase plate. We verify the feasibility of generating vortex beams by simulation analysis, and analyze vortex beams generated by the designed structure and the general spiral phase plate structure by simulation.

The Fabry-Perot interferometer (FP) can be used as a kind of filter for obtaining spectral information of targets in several wavelength ranges such as in the visible or infrared regions. So far, the spectral imaging devices based on FP effect mainly include the electrically controlled liquid-crystal filtering structures and the micro-electro-mechanical filtering architectures (MEMS). MEMS are generally micro-structures that integrate micro-sensors for converting incident microbeams into arrayed electronic signals and micro-actuators. The MEMS-FP filter constructed by combining the MEMS and FP functions, can be further integrated into a chip-level imaging spectrometer to achieve spectral imaging operation. The design of distributed Bragg reflectors (DBRs) is an important part to obtain a high transmittance for MEMS-FP structure. Different number of layers of optical film is calculated and compared in this paper and the transmittance can reach 82% and the FWHM is ~ 1nm in the infrared region of 3-5um. Angle of incidence is also considered and the simulation result shows poor robustness. We propose that two liquid-crystal microlens arrays can be mounted on FP arrays to get a high filling-factor and a more flexible range of incident angles.

Light vector polarization as a fundamental property of lightwave, can be used to effectively distinguish objects in complicated circumstance including surface shape and materials type and transmission medium. As shown, polarization imaging is an advanced information acquirement method which combines the light intensity image and light vector vibration behaviors, which is the direction of electric field of incident lightwaves. A typical microgrid polarimeter with a minimum repeat unit is composed of four pixelated linear polarizer demonstrating different vibration directions. Compared with full polarization information, the polarization image obtained has only one quarter polarization information in each direction. Thus, it will influence the accuracy of other information such as Stokes components, the degree of linear polarization (DoLP), and the angle of polarization (AoP). In this paper we propose a polarization demosaicing network to address the poloarized image demosaicing issue, which are then recovered into the original polarized image. This network aims to improve the accuracy of DoLP and AoP of the targets by adjusting three Stokes components of the network output. We already remove the batch normalization (BN) commonly used in CNN, and thus use a customized loss function to make it suitable for polarization image demosaicing. The experimental results show that network has demonstrated a best peak signal-to-noise ratio (PSNR) and then richer image detail and polarization target information than that of the original image.

As demonstrated, surface plasmons (SPs) stimulated by incident lightwaves are one of the most popular research fields, currently. The researches about the remarkable generation and efficient transmission and effective manipulation of relatively strong SPs are generally limited by a range constraint of wavelength or sub-wavelength-scaled structures. So far, the interaction between the electromagnetic field and the free electrons over the metal and medium interface or special metal micro-nano-structure has been mainly studied. In reality, a type of adjustable ionic exciter device is needed, which lead to a new focus about the adjustable ionic exciter materials. At present, two-dimensional graphene materials already demonstrate several excellent optical and electrical properties, and their conductivity and dielectric constant can be easily affected by external bias electric field, so as to exhibit a prospect as a kind of basic materials for adjustable and other excitation components. In this paper, the adjustable properties of single crystal graphene are studied. The effects based on the factors including the temperature and the scattering rate and the chemical potential corresponding to some parameters such as the conductivity and dielectric constant of graphene are analyzed carefully. In addition, the composite structure of the graphene grating nano-apexes is designed, which is characterized based on the multi-frequency points resonance according to incident light at the waveband of 4~11μm. The key graphene-based structure is modeled and simulated by the FDTD solution based on a finite difference time domain method under the different chemical potential. Then, the transmission and reflection and absorption behaviors of the graphene-based structure were analyzed according to the near electric field intensity distribution curves.

Currently, optical antenna has already become a research hotspot because of its remarkable local field enhancement effect and resonance propagation characteristics. Optical antennas are usually designed as a type of sub-wavelength-scaled metal structures. By transmitting the field enhancement signal of the optical antenna to the infrared detector, the weak signal detection ability of the infrared detector can be improved. As demonstrated, the resonant wavelength of the local surface plasmons is determined by the structure and material of the antenna and also the material properties of the surrounding medium. By changing the geometry of the antenna or the dielectric characteristics of the circumstance medium, the response frequency of the optical antenna can be regulated. As a two-dimensional material with unique electrical and optical properties, the dielectric properties of graphene can be regulated by applied bias voltage. By selecting the geometry of the antenna and applying bias voltage, the optical antenna with unique characteristics can be obtained. In this paper, an optical antenna with a graphene-silica-silicon trilayer structure is designed and a planar-tip array is fabricated over the graphene layer. The influence of the geometry of the planar-apex array and the thickness of the silica dielectric layer corresponding to the optical properties of the graphene antenna are analyzed. Simulation results show that by changing the shape of the planar-tip and the thickness of the silica dielectric layer, the position and intensity of the absorption peak of the graphene optical antenna can be controlled effectively. At the same time, under the control of external bias voltage, the resonance peak also appears an obvious movement of a maximum range of about three microns.

Surface plasmons, as a local electromagnetic field mode generated or stimulated at the interface between common metal and dielectric, can be used to greatly break through the optical diffraction limit and also localize the electric-field and then light energy in a sub-wavelength scale. It is already a research hotspot in recent years. As shown, several patterned metal nano-array can be utilized to produce relatively strong surface plasmon resonance, so as to achieve a nano-scaled localized light field on the surface of the functioned metal structure. In this paper, silicon dioxide materials are used as the substrate, and the common gold materials are fabricated into a metal film, and then the sub-wavelength metal nano-tip arrays with several morphology such as the cone-shaped, the triangular-pyramid-shaped, and the quadrangular-pyramid-shaped, are designed respectively. The functioned metal nano-structures are symmetric and asymmetric coating mode. The electric field distribution characteristics of the structure under the internal excitation mode of the incident light with vertical incidence are analyzed. The simulations show that the local field enhancement can be clearly observed at the nano-tip of the cone-shaped in the asymmetric case, but the symmetry is not. Analysis shows that the destructive interference occurs when the surface plasmons are excited by a linearly polarized light on both sides of a conical structure propagate to its top, so failure to produce focusing effect. Therefore, to the case of symmetrical film through adjusting the incident angle of light, different incident angles will affect the enhancement of the local field at the tip of the cone.

This paper proposes a new irregular remote sensing object detection algorithm that different from the ROI or rotating BOX obtained by traditional one. The architecture is designed to jointly learn four bounding box corner points and their association via two branches of the same sequential prediction process. The algorithm predicts four key points of the object and their associated connection, Bounding Box Fields(BBF) via convolutional neural network(CNN), and thus obtains the detail spatial distribution of the objects.

In order to improve the positioning accuracy of the key points, network architecture reduced Receptive Field from large to small stage by stage. It has achieved ROI free finally. In this method, the object detection problem is framed as CNN convolution point detection and bounding box field detection, it achieved the one stage object detection with high precision and high speed.

We verified the effectiveness and efficiency of the algorithm through experiments, which proved that the new data structure could locate the object attitude and spatial direction more accurately in real time with strong practicability.

At present, the restoration of turbulent degraded images is a worldwide problem in the fields of astronomical imaging. Atmospheric turbulence is the reason why images will be blurred, which gravely interferes the object of recognition and the detection of images. This paper presents a restoration method for the turbulent degraded images based on the image saliency edge selection and the L₀ norm constraint, which aims to recover sharp images from the turbulent degraded images. The proposed method imposes the L₀ norm sparse constraint on the latent image, and uses the method of split Bregman to solve the problem of optimization. To avoid the influence of the tiny details on the point spread function (PSF) , we use the image saliency algorithm to build a weighted model to select salient edges from the latent images. Based on the salient edge in the gradient domain, the proposed method establishes an estimation model of the point spread function. The calculation part of the point spread function is solved accurately by using the fast Fourier transform (FFT) in the frequency domain. The proposed method uses the multi-scale pyramid strategy to alternatively solve the point spread function and the latent images, which can obtain the final accuracy of the point spread function.

Attention mechanism in deep learning is similar to information selection mechanism, and the goal of attention is to select critical information for the current task. In hyperspectral classification, the distinction of some categories depends on the subtle differences, however, most of the classification methods have the problem of insufficient expression ability to discriminate the fine differences of categories. In this paper, a classification method based on group attention is proposed to enhance the difference of hyperspectral data between categories. Firstly, we slice the hyperspectral sample into several groups on spectral channels, and extract the group CNN features. Then we use the attention module to obtain the attention weights for each spectral group. Finally, the "feature recalibration" strategy is used to recalibrate the spectral group CNN features. The experiment show that the proposed approach can improve the classification accuracy of categories with subtle differences.

Stripe is a common degradation phenomenon in remote sensing images. The variation-based de-striping method, due to the defect of the model itself, always has an unnecessary influence on the stripe-free area while correcting the stripe, and cannot satisfy some requirements in high-precision quantitative applications or sensitive data processing of remote sensing images. This paper proposes a high-precision stripe correction method, which first detects the position of the stripes, and then uses the interpolation idea to correct the stripe to solve the fidelity problem of the stripe-free area in the de-striping process. We use the rational assumption that the derivative of the real signal in the stripe region (to be repaired) is consistent with the derivative of the observed signal, and then selects cubic Hermite spline interpolation method for de-striping, which can uses the derivative information of the region to be repaired (ie, the derivative information of the stripe region) to overcoming the difficulty of the existing interpolation de-stripe method not being able to work well when the stripes is too wide. The experimental results show that our method can effectively remove the stripes and maintain the stripe-free area intact.

Land cover composition and change are important aspects for many scientific research and socioeconomic assessments. The multi-date land cover change detection is generally more difficult and time-consuming to select enough training samples when considering multi-date image labels at the same location. To improve the accuracy for multi-date change detection, this study proposed a new algorithm framework, combining self-learning and relearning algorithm. Wuhan was selected as the experimental area, and Landsat images in 2005 and 2016 were used to extract six main types of change classes. Firstly, PCM (primitive co-occurrence matrix) and the minimum class certainty are used to ensure the high confidence of selected candidate set samples, and then the most informative samples are identified for classification from the candidate samples. To save computing costs, we adopt clustering method to reduce the self-relearning samples. Based on our experimental results, the self-relearning algorithm increases the final classification accuracy by approximately 2.5% (from 92.64% to 95.09%) in the case of using few initial training samples, providing a feasible solution for the multi-date change detection.

Because the structure of the moving target in the air is complex and its composition is not single, it is impossible to simply regard the measured spectral signal as the sum of all parts. For spectral acquisition of complex small target in salient region, a target saliency region unmixing method is innovatively proposed in this paper. The abundance information of sky background, target region and salient regions was obtained by analyzing the infrared image. According to the measured spectra and their weighted linear combination relationship, the precise spectral signal of the target saliency region can be estimated. The experimental results show that, through this method, the spectra of each salient regions on the target is more accurate and have more abundant details. The comprehensive judgment result of each salient regions spectra can effectively improve the target recognition accuracy.

In this paper, we introduce a novel nonuniformity correction (NUC) algorithm for infrared focal-plane array (IRFPA). It is based on layers technique. First, the Rolling Guidance Filter (RGF) is utilized to decompose the raw IR image into a low frequency part and a high frequency part. Then, an adaptive temporal high-pass filter is utilized to filter the high frequency part by making use of the gradient and amplitude of it to estimate the Fixed Pattern Noise (FPN). The proposed scheme use the frames with large displacement to estimate the FPN to alleviate the ghosting artifacts in case of scene moves slowly. At Last, the estimated FPN is subtracted from the pristine image to obtain the correct result. Experiments with synthetic and real IR video demonstrate that the proposed method has better NUC performance and less artifacts than the state-of-the-arts.

Images often suffer from low visibility under nonuniform illumination, weak luminance and backlight environment. This paper describes a novel approach to improvement the visualization of poor light conditions. Firstly, we raise the global brightness using an adaptive exponent induced function. To enhance the local detail perception, the local contrast is boosted by contrast preserving which utilizes human vision system model. To not bias from original image, we generate the contrast combined original image and global illuminance enhance output in the gradient domain. To reduce artifacts, the guided filter is employed to estimate the local mean illuminance when transform the contrast. The experimental results demonstrate that our proposed method has a pleasant visual effect and low computational complexity than the state of the arts.

Speckle noise limits the usage of synthetic aperture radar (SAR) for object recognition and segmentation tasks. Most traditional methods sacrifice useful image information to achieve speckle reduction. The classic method based on local sliding window filtering has obvious side effect of erasing object edges and blurring texture information comparing with ground truth image. Another widely used method is convolutional neural network based on mean squared error, the visual effect of denoised image is not satisfactory even though MSE loss can have higher peak signal-to-noise ratio (PSNR) performance. In this paper, we present a cascade network to address this problem, namely SRNet, which employs an asymmetric architecture for the task of speckle noise reduction. The cascade architecture can supervise the network to revise on both pixel-wise level and feature-wise level by calculating correlation coefficient loss on the feature maps. In the meanwhile, we utilize the auxiliary loss on the intermediate results to accelerate the convergence of the network. The proposed network preserves the edge texture details much better than other compared methods.

The hyperspectral imaging, adding many dimensions, has practical significance for robust face recognition. However, for hyperspectral face recognition, the main problems are small sample collection, low signal-to-noise ratio and inter band misalignment. In view of these problems, we propose a hyperspectral face recognition method based on SLRC (superposed linear representation classifier) for single sample problem. In the proposed method, one sample from each class is selected as training data, then the rest samples as test data. Since hyperspectral images have multiple bands, we average all bands as prototype dictionaries, and the difference between each band and the corresponding prototype dictionary as variation dictionary. Compared with other sparse representation classification methods, the proposed method can directly use a single sample to train hyperspectral face recognition and has no handcraft feature extraction. Experiments on the hyperspectral face database (PloyU-HSFD) validate that the proposed method can not only greatly increase the accuracy in single sample hyperspectral face recognition, but also improve the computation speed.

This paper proposes the generation of a pedestrian ROI region, which is mainly aimed at pedestrian segmentation in far-infrared (FIR) images of in-vehicle systems. Since the FIR image is a grayscale image, the pixel value of the pedestrian is usually higher than the background, so the previous segmentation method is mainly threshold segmentation. However, this method will cause problems due to the uneven brightness of pedestrians caused by pedestrian wear, etc. We propose a new method for generating pedestrian ROI regions, which is based on the combination of image region merging and pixel-intensity vertical projection, and adopts the time domain semantic model to constrain the parameter space. Experiments show that our method has achieved good results in urban scenes.

Based on the imaging mechanism of infrared image, we carry out the research of infrared spectral image inversion method for expanding data sets. Firstly, the radiation brightness of the long-wave infrared image is inverted to temperature, and then the radiation brightness of the medium-wave infrared image is calculated by using the inverted temperature. In order to improve the accuracy of infrared image inversion, the influence of range radiation and atmospheric transmittance on infrared radiation will be removed in the calculation process. Finally, we evaluate the effectiveness of this method.

This paper is mainly aimed at the development requirement of infrared seeker precision guided weapon system. The infrared sensor of seeker is modeled and simulated, and the image sequence is output. The imaging simulation quality of infrared seeker sensor and the adaptability of terminal guidance algorithm based on simulation real-time graph are evaluated from the perspectives of multi-granularity similarity parameter index system construction, multi-saliency fusion feature consistency and bio-visual feature similarity evaluation. At the same time, the analytic hierarchy process (AHP) model is applied to the process of similarity evaluation to solve the comprehensive decision-making problem in multi-index similarity evaluation, and to evaluate the performance of the simulation system from the perspective of the combination of human-computer intelligence, so as to guide the improvement of the simulation system. Based on the above research results, a set of mature and reliable simulation similarity evaluation software is formed. The research results are of great significance to the development and standardization of the whole infrared simulation technology.