SHEN Y , LI J , LIN W F , et al . Camouflaged target detection based on snapshot multispectral imaging [J ] . Remote Sensing , 2021 , 13 ( 19 ): 3949 .

甘鑫 , 高欣健 , 钟彬彬 , 等 . 基于有限样本的大气偏振模式生成方法 [J ] . 光电工程 , 2021 , 48 ( 5 ): 200331 .

GAN X , GAO X J , ZHONG B B , et al . A few-shot learning based generative method for atmospheric polarization modelling [J ] . Opto-Electronic Engineering , 2021 , 48 ( 5 ): 200331 . (in Chinese)

KARIM S , TONG G , LI J , et al . Current advances and future perspectives of image fusion: a comprehensive review [J ] . Information Fusion , 2023 , 90 : 185 - 217 .

SATTAR S , LAPRAY P , FOULONNEAU A . Review of spectral and polarization imaging systems [J ] . Proceedings of SPIE , 2020 , 11351 : 113511Q.

WANG X , ZHANG Y H , MA X , et al . Compressive spectral imaging system based on liquid crystal tunable filter [J ] . Optics Express , 2018 , 26 ( 19 ): 25226 - 25243 . DOI: 10.1364/OE.26.025226 http://doi.org/10.1364/OE.26.025226 Liquid crystal tunable filters (LCTF) are extensively used in hyperspectral imaging systems to successively acquire different spectral components of scenes by adjusting the center wavelength of the filter. However, the spectral and spatial resolutions of the imager are limited by the bandwidth of LCTF, and the pitch dimension of the detector, respectively. This paper applies compressive sensing principles to improve both of the spatial and spectral resolutions of the LCTF-based hyperspectral imaging system. An accurate transmission model of the LCTF is used to represent its bandpass filtering effects on the spectra. In addition, a random coded aperture placed behind the LCTF is used to modulate the spectral images in the spatial domain. Then, the three-dimensional encoded spectral images are projected onto a two-dimensional detector. Benefiting from the spectral-dependent transmission property of the LCTF, information of the entire spectrum is collected by a few snapshots using different center wavelengths of the LCTF. Super-resolution hyperspectral images can be reconstructed from a small set of compressive measurements by solving a convex optimization problem. Simulations and experiments show that the proposed method can effectively improve the spectral and spatial resolutions of traditional LCTF-based spectral imager without changing the structures of the LCTF and detector. Finally, a multi-channel spectral coding method is proposed to further increase the compression capacity of the system.

ZHOU P W , WANG Y , HE Z E , et al . Blurred spectral images restoration technology for AOTF imaging spectrometer based on dual-path architecture [J ] . Optics Express , 2022 , 30 ( 12 ): 21746 - 21757 . DOI: 10.1364/OE.459972 http://doi.org/10.1364/OE.459972 The acousto-optic tunable filter (AOTF) imaging spectrometer is a staring-type instrument that acquires spectral images of different bands asynchronously and sequentially, and is susceptible to platform jitter and target motion. When the integration time of the detector increases, serious spectral images degradation occurs along with a large band-to-band misregistration. The dual-path optical configuration, capable of simultaneously acquiring diffracted and undiffracted beams of AOTF, has been shown to be effective in solving the problem of band-to band misregistration. Hence by solving the remaining spectral image degradation problems, the application range of AOTF will transcend the previous limitation. Therefore, this paper presents a new, to the best of our knowledge, method of capturing a series of short-exposure undiffracted beam images to calculate the blur kernel of the diffracted spectral images, since the high-intensity characteristics of the undiffracted beam can also be exploited in the dual-path system. The paper starts with analyzing the impact of the platform on spectral images, and then demonstrates the development and calibration of a dual-path based spectral image deblurring method. Finally, laboratory experiment verifies that the resulting blur kernel can be effectively used for spectral image restoration, thus demonstrating the robustness of this technique.

李虎 , 刘雪峰 , 姚旭日 , 等 . 并行压缩感知计算层析成像光谱 [J ] . 光谱学与光谱分析 , 2023 , 43 ( 2 ): 348 - 355 .

LI H , LIU X F , YAO X R , et al . Blockcompressed sensing computed tomography imaging spectrometry [J ] . Spectroscopy and Spectral Analysis , 2023 , 43 ( 2 ): 348 - 355 . (in Chinese)

ZHANG Y , WANG H , SUN J H , et al . Real-time polarization spectral detection based on a dual-dispersion and coded-aperture imaging system [J ] . IEEE Transactions on Geoscience and Remote Sensing , 2023 , 61 : 5602412 .

SATTAR S , LAPRAY P , AKSAS L , et al . Snapshot spectropolarimetric imaging using a pair of filter array cameras [J ] . Optical Engineering , 2022 , 61 ( 4 ): 043104 .

聂晓风 , 唐林波 , 张国阳 , 等 . 基于偏振信息的低可探测目标检测方法 [C ] // 第十五届全国信号和智能信息处理与应用学术会议论文集 . 重庆 : 中国高科技产业化研究会智能信息处理产业化分会 , 2022 : 220 - 225 .

NIE X F , TANG L B , ZHANG G Y , et al . Low detectable target detection based on polarization information [C ] // Proceedings of the 15th National Conference on Signal and Intelligent Information Processing and Application . Chongqing : Intelligent Information Processing Industrialization Branch , China High-tech Industrialization Association, 2022 : 220 - 225 . (in Chinese)

YANG X L , ZHAO M , SHI S , et al . Deep constrained energy minimization for hyperspectral target detection [J ] . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2022 , 15 : 8049 - 8063 .

CHANG C , CAO H J , SONG M P . Orthogonal subspace projection target detector for hyperspectral anomaly detection [J ] . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2021 , 14 : 4915 - 4932 .

杨威 , 王晓曼 , 赵海丽 , 等 . 一种改进的水下偏振图像融合算法研究 [J ] . 长春理工大学学报(自然科学版) , 2021 , 44 ( 4 ): 43 - 49 .

YANG W , WANG X M , ZHAO H L , et al . Research on an improved underwaterpolarization image fusion algorithm [J ] . Journal of Changchun University of Science and Technology (Natural Science Edition) , 2021 , 44 ( 4 ): 43 - 49 . (in Chinese)

SHEN Y , LIN W F , WANG Z F , et al . Rapid detection of camouflaged artificial target based on polarization imaging and deep learning [J ] . IEEE Photonics Journal , 2021 , 13 ( 4 ): 7800309 .

REN S , HE K , GIRSHICK R , et al . Faster R-CNN: towards real-time object detection with region proposal networks [J ] . IEEE Transactions on Pattern Analysis and Machine Intelligence , 2017 , 39 ( 6 ): 1137 - 1149 . DOI: 10.1109/TPAMI.2016.2577031 http://doi.org/10.1109/TPAMI.2016.2577031 State-of-the-art object detection networks depend on region proposal algorithms to hypothesize object locations. Advances like SPPnet [1] and Fast R-CNN [2] have reduced the running time of these detection networks, exposing region proposal computation as a bottleneck. In this work, we introduce a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals. An RPN is a fully convolutional network that simultaneously predicts object bounds and objectness scores at each position. The RPN is trained end-to-end to generate high-quality region proposals, which are used by Fast R-CNN for detection. We further merge RPN and Fast R-CNN into a single network by sharing their convolutional features-using the recently popular terminology of neural networks with 'attention' mechanisms, the RPN component tells the unified network where to look. For the very deep VGG-16 model [3], our detection system has a frame rate of 5 fps (including all steps) on a GPU, while achieving state-of-the-art object detection accuracy on PASCAL VOC 2007, 2012, and MS COCO datasets with only 300 proposals per image. In ILSVRC and COCO 2015 competitions, Faster R-CNN and RPN are the foundations of the 1st-place winning entries in several tracks. Code has been made publicly available.

WANG C Y , BOCHKOVSKIY A , LIAO H Y M . YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors [C ] // Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver, BC, Canada:IEEE , 2023 : 7464 - 7475 .

HUBER-LERNER M , HADAR O , ROTMAN S R , et al . Hyperspectral band selection for anomaly detection: the role of data gaussianity [J ] . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2016 , 9 ( 2 ): 732 - 743 .

王霞 , 苏子航 , 赵家碧 , 等 . 基于交织序列的偏振方向映射方法 [J ] . 应用光学 , 2021 , 42 ( 5 ): 877 - 883 . DOI: 10.5768/JAO202142.0503001 http://doi.org/10.5768/JAO202142.0503001 偏振方向是偏振成像技术中的基本信息之一，能够反映出场景物体表面的方向信息，因此场景中偏振方向的分布对场景信息的解译有着重要的作用。传统的偏振方向表示方法存在着二维或三维信息映射方法的映射结果维度较高和由映射关系的距离相关性低而引起的映射方法噪声鲁棒性低问题。提出了具有距离保持特性的基于交织序列的偏振方向信息一维数值映射方法。选择从斯托克斯矢量分量中提取方向信息作为偏振方向的二维映射方法，再基于交织序列方法将二维数据映射为一维数值；之后研究了距离保持特性与映射方法对噪声鲁棒性的关系，并分析了该文所提出的偏振方向映射方法与传统偏振角映射方法的距离保持特性；进行了仿真实验与真实场景实验。仿真实验中，该文所提方法映射结果的PSNR均值相较于偏振映射方法提升了6%；真实场景实验中，该文方法映射结果相较于斯托克斯矢量映射方法降低了信息维度，相较于偏角映射方法，NMF特征的NRPSNR均值提升约14%，证明了所提出的映射方法具有低信息维度及较好的噪声鲁棒性。

WANG X , SU Z H , ZHAO J B , et al . Sequence-interleaving mapping method for direction of polarization [J ] . Journal of Applied Optics , 2021 , 42 ( 5 ): 877 - 883 . (in Chinese) DOI: 10.5768/JAO202142.0503001 http://doi.org/10.5768/JAO202142.0503001 Direction of polarization, as one of the basic information in polarization imaging technology, can reflect the direction information of object surface in the scene. Therefore, the distribution of polarization direction in the scene plays an important role in the interpretation of the scene information. The conventional representation method of polarization direction has two main problems: one is high dimension of mapping results for 2-D or 3-D information mapping methods, another one is low noise-robustness of mapping methods caused by the low distance correlation of mapping relationship. A 1-D numerical value mapping method of polarization direction information based on the interleaved sequence with distance-preserving property was proposed. Firstly, the direction information was extracted from the stokes vector component as the 2-D mapping method of polarization direction. Based on the interleaved sequence method, the 2-D numerical value was mapped into the 1-D numerical value. Then, the relationship between the distance-preserving property and mapping method to noise-robustness was studied, and the distance-preserving property of the proposed mapping method of polarization direction and the conventional mapping method of polarization angle was analyzed. Finally, the simulation experiment and real-scene experiment were conducted. In the simulation experiment, the peak signal to noise ratio (PSNR) mean of the mapping results of proposed method was increased by 6% compared to the mapping method of polarization. In the real-scene experiment, the mapping results of proposed method reduced the information dimension compared to the mapping method of stokes vector and the no-reference PSNR (NRPSNR) mean of the non-negative matrix factorization (NMF) features was increased by 14% compared to the mapping method of polarization angle, which proved that the proposed mapping method has low information dimension and good noise-robustness.

XU H , WANG Y , WU Y J , et al . Infrared and multi-type images fusion algorithm based on contrast pyramid transform [J ] . Infrared Physics & Technology , 2016 , 78 : 133 - 146 .

WAN M J , GU G H , QIAN W X , et al . Stokes-vector-based polarimetric imaging system for adaptive target/background contrast enhancement [J ] . Applied Optics , 2016 , 55 ( 21 ): 5513 - 5519 .

高新波 , 莫梦竟成 , 汪海涛 , 等 . 小目标检测研究进展 [J ] . 数据采集与处理 , 2021 , 3 ( 36 ): 391 - 417 .

GAO X B , MO M J C , WANG H T,etal . Recent advances in small object detection [J ] . Journal of Data Acquisition and Processing , 2021 , 3 ( 36 ): 391 - 417 . (in Chinese)

ZHANG Y , ZHANG Y , ZHAO H J , et al . Improved atmospheric effects elimination method for pBRDF models of painted surfaces [J ] . Optics Express , 2017 , 25 ( 14 ): 16458 - 16475 . DOI: 10.1364/OE.25.016458 http://doi.org/10.1364/OE.25.016458 A method for eliminating atmospheric effects in polarimetric imaging remote sensing detection was developed by combining the shadowing method and radiative transfer (RT) model. First, a polarized bidirectional reflectance distribution function (pBRDF) model of painted surfaces was constructed. Using the resulting polarimetric radiance composition, the atmospheric effects elimination method was developed and compared to Shell's method. Experiments were performed using a liquid-crystal-variable-retarder-based imaging polarimeter to obtain the surface pBRDFs. The proposed method showed better performance under different weather conditions than Shell's method. Furthermore, the error was below 4.8% in the proposed method (6.8% in Shell's method), indicating improved quantitative accuracy of the target physical parameters in remote sensing.

ZHANG Y , WANG H , LI H S , et al . Optimization model of signal-to-noise ratio for a typical polarization multispectral imaging remote sensor [J ] . Sensors , 2022 , 22 ( 17 ): 6624 .

CHEN G , WANG H T , CHEN K , et al . A survey of the four pillars for small object detection: multiscale representation, contextual information, super-resolution, and region proposal [J ] . IEEE Transactions on Systems, Man, and Cybernetics: Systems , 2022 , 52 ( 2 ): 936 - 953 .

沈英 , 刘贤财 , 王舒 , 等 . 基于偏振编码图像的低空伪装目标实时检测 [J ] . 兵工学报 , 2024 , 45 ( 5 ): 1374 - 1383 . DOI: 10.12382/bgxb.2022.1289 http://doi.org/10.12382/bgxb.2022.1289 偏振可以提高无人机的自主侦察能力,但易受到探测角度和目标材质的影响,从而降低偏振检测的鲁棒性。为此,提出一种基于偏振图像的低空伪装目标实时检测算法YOLO-P,采用融合多偏振方向信息的编码图像作为输入,应用三维卷积模块提取不同偏振方向图像之间的联系特征;引入特征增强模块对多层次特征进行进一步增强;采用跨层级特征聚合网络,充分利用不同尺度的特征信息,完成特征的有效聚合,最终联合多通道特征信息输出检测结果。构建包含10类目标的低空伪装目标偏振图像数据集PICO(Polarization Image of Camouflaged Objects)。在PICO数据集上的实验结果表明,新方法可以有效检测伪装目标,mAP 0.5:0.95 达到52.0%,mAP 0.5 达到91.5%,检测速率达到55.0帧/s,满足实时性要求。

SHEN Y , LIU X C , WANG S , et al . Real-time detection of low-altitude camouflage targets based on polarization coded images [J ] . Acta Armamentarii , 2024 , 45 ( 5 ): 1374 - 1383 . (in Chinese) DOI: 10.12382/bgxb.2022.1289 http://doi.org/10.12382/bgxb.2022.1289 Polarization can improve the autonomous reconnaissance capability of unmanned aerial vehicle, but it is easily interfered by the variation of detection angle and target materials, which affects the robustness of polarization detection. In this paper, a real-time low-altitude camouflaged target detection algorithm of YOLO-Polarization based on polarized images is proposed. The coded image fused with multi-polarization direction information is used as input, the 3D convolution module is applied to extract the connection features from the different polarization direction images, and a feature enhancement module (FEM) is introduced to further enhance the multi-level features. In addition, the cross-level feature aggregation network is adopted to make full use of the feature information of different scales to complete the effective aggregation of features, and finally combined with multi-channel feature information output detection results. A dataset consisting of polarized images of low-altitude camouflaged targets (PICO) which include 10 types of targets is constructed. The experimental results based on PICO dataset show that the proposed method can effectively detect the camouflaged targets, with mAP 0.5:0.95 up to 52.0% and mAP 0.5 up to 91.5%. The detection rate achieves 55.0 frames/s, which meets the requirement of real-time detection.