2021
2021
-
Record 37 of
Title:Tilt Correction Toward Building Detection of Remote Sensing Images
Author(s):Liu, Kang(1,5); Jiang, Zhiyu(2); Xu, Mingliang(3); Perc, Matjaz(4); Li, Xuelong(2)Source: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Volume: 14 Issue: DOI: 10.1109/JSTARS.2021.3083481 Published: 2021Abstract:Building detection is a crucial task in the field of remote sensing, which can facilitate urban construction planning, disaster survey, and emergency landing. However, for large-size remote sensing images, the great majority of existing works have ignored the image tilt problem. This problem can result in partitioning buildings into separately oblique parts when the large-size images are partitioned. This is not beneficial to preserve semantic completeness of the building objects. Motivated by the above fact, we first propose a framework for detecting objects in a large-size image, particularly for building detection. The framework mainly consists of two phases. In the first phase, we particularly propose a tilt correction (TC) algorithm, which contains three steps: texture mapping, tilt angle assessment, and image rotation. In the second phase, building detection is performed with object detectors, especially deep-neural-network-based methods. Last but not least, the detection results will be inversely mapped to the original large-size image. Furthermore, a challenging dataset named Aerial Image Building Detection is contributed for the public research. To evaluate the TC method, we also define an evaluation metric to compute the cost of building partition. The experimental results demonstrate the effects of the proposed method for building detection. © 2008-2012 IEEE.Accession Number: 20212310459709 -
Record 38 of
Title:Optical neuromorphic processing based on Kerr microcombs
Author(s):Xu, X.(1,2); Tan, M.(1); Wu, J.(1); Boes, A.(3); Corcoran, B.(2); Nguyen, T.(3); Chu, S.T.(4); Little, B.E.(5); Morandotti, R.(6,7); Mitchell, A.(3); Hicks, D.(1,8); Moss, D.J.(1)Source: Optics InfoBase Conference Papers Volume: Issue: DOI: null Published: 2021Abstract:We report a new approach to ONNs based on integrated Kerr micro-combs that is programmable, highly scalable and capable of reaching ultra-high speeds. We demonstrate a single neuron perceptron at 11.9 Giga-OPS at 8 bits per OP, or 95.2 Gbps. We then demonstrate a convolutional accelerator operating beyond 11 TeraOPs/s. We test the perceptron on handwritten-digit recognition and cancer-cell detection - achieving over 90% and 85% accuracy, respectively. © OSA 2021, © 2021 The Author(s)Accession Number: 20214811250748 -
Record 39 of
Title:Tunable wettability pattern transfer photothermally achieved on zinc with microholes fabricated by femtosecond laser
Author(s):Li, Fengping(1,2); Feng, Guang(2); Yang, Xiaojun(3); Lu, Chengji(2); Ma, Guang(2); Li, Xiaogang(2); Xue, Wei(2); Sun, Haoran(2)Source: Micromachines Volume: 12 Issue: 5 DOI: 10.3390/mi12050547 Published: 2021Abstract:A quickly tunable wettability pattern plays an important role in regulating the surface behavior of liquids. Light irradiation can effectively control the pattern to achieve a specific wettability pattern on the photoresponsive material. However, metal oxide materials based on light adjustable wettability have a low regulation efficiency. In this paper, zinc (Zn) superhydrophobic surfaces can be obtained by femtosecond-laser-ablated microholes. Owing to ultraviolet (UV) irradiation increasing the surface energy of Zn and heating water temperature decreasing the surface energy of water, the wettability of Zn can be quickly tuned photothermally. Then, the Zn superhydrophobic surfaces can be restored by heating in the dark. Moreover, by tuning the pattern of UV irradiation, a specific wettability pattern can be transferred by the Zn microholes, which has a potential application value in the field of new location-controlled micro-/nanofluidic devices, such as microreactors and lab-on-chip devices. © 2021 by the authors. Licensee MDPI, Basel, Switzerland.Accession Number: 20212210439340 -
Record 40 of
Title:Single perceptron at 12 GigaOPs based on a microcomb for versatile, high-speed scalable, optical neural networks
Author(s):Tan, M.(1); Xu, X.(2); Wu, J.(1); Boes, A.(3); Corcoran, B.(2); Nguyen, T.G.(3); Chu, S.T.(4); Little, B.E.(5); Morandotti, R.(6,7); Mitchell, A.(3); Hicks, D.G.(1,8); Moss, D.J.(1)Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 11690 Issue: DOI: 10.1117/12.2584011 Published: 2021Abstract:Optical artificial neural networks (ONNs) have significant potential for ultra-high computing speed and energy efficiency. We report a novel approach to ONNs that uses integrated Kerr optical micro-combs. This approach is programmable and scalable and is capable of reaching ultra-high speeds. We demonstrate the basic building block ONNs- A single neuron perceptron-by mapping synapses onto 49 wavelengths to achieve an operating speed of 11.9 x 109 operations per second, or Giga-OPS, at 8 bits per operation, which equates to 95.2 gigabits/s (Gbps). We test the perceptron on handwritten-digit recognition and cancer-cell detection- A chieving over 90% and 85% accuracy, respectively. By scaling the perceptron to a deep learning network using off-the-shelf telecom technology we can achieve high throughput operation for matrix multiplication for real-time massive data processing. © 2021 SPIE.Accession Number: 20212110381609 -
Record 41 of
Title:High-speed focusing and scanning light through a multimode fiber based on binary phase-only spatial light modulation
Author(s):Geng, Yi(1,2); Chen, Hui(1,2); Zhang, Zaikun(1,2); Zhuang, Bin(1,2); Guo, Haitao(1,2); He, Zhengquan(1); Kong, Depeng(1)Source: Applied Physics B: Lasers and Optics Volume: 127 Issue: 2 DOI: 10.1007/s00340-021-07573-1 Published: February 2021Abstract:A binary phase-only modulation technique was proposed to focus and scan light through a multimode fiber (MMF) based on spatial light modulator (SLM). For the same number of modulation modes, the number of iterations using this method is only 1/256 of that using phase-only iterative optimization or 1/3 of that using phase-only computation optimization, and the modulation time is at least one to two orders of magnitude shorter than previous wavefront shaping systems. Focusing and scanning light through an MMF with a 105 μm core diameter and 5 m length was achieved experimentally. This method can be extended to focus and scan light at multiple planes along the axial direction by modifying the input wavefront accordingly. © 2021, The Author(s), under exclusive licence to Springer-Verlag GmbH, DE part of Springer Nature.Accession Number: 20210509853416 -
Record 42 of
Title:Photonic convolutional accelerator and neural network in the Tera-OPs regime based on Kerr microcombs
Author(s):Xu, Xingyuan(1); Tan, Mengxi(1); Corcoran, Bill(2); Wu, Jiayang(1); Boes, Andreas(3); Nguyen, Thach G.(3); Chu, Sai T.(4); Little, Brent E.(5); Hicks, Damien G.(1,6); Morandotti, Roberto(7,8); Mitchell, Arnan(3); Moss, David J.(1)Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 11689 Issue: DOI: 10.1117/12.2584017 Published: 2021Abstract:We report a photonic RF arbitrary waveform generator and phase encoded signal generator using a soliton crystal micro-comb source with a free spectral range (FSR) of 48.9 GHz, that provides 80 wavelengths. We achieve arbitrary waveform shapes including square waveforms with a tunable duty ratio from 10% to 90%, sawtooth waveforms with a slope ratio of 0.2 to 1, and a symmetric concave quadratic chirp waveform with an instantaneous frequency of sub GHz. The phase encoded signal generator achieves a high pulse compression ratio of 30 and phase encoding rates from 2 to 6 Gb/s This work verifies the effectiveness of using microcombs for high-performance, broad bandwidth, nearly user-defined RF waveform generation. © 2021 SPIE.Accession Number: 20212110380609 -
Record 43 of
Title:New results on small and dim infrared target detection
Author(s):Wang, Hao(1); Zhao, Zehao(1,2); Kwan, Chiman(3); Zhou, Geqiang(4); Chen, Yaohong(1)Source: Sensors Volume: 21 Issue: 22 DOI: 10.3390/s21227746 Published: November-2 2021Abstract:Real-time small infrared (IR) target detection is critical to the performance of the situa-tional awareness system in high-altitude aircraft. However, current IR target detection systems are generally hardware-unfriendly and have difficulty in achieving a robust performance in datasets with clouds occupying a large proportion of the image background. In this paper, we present new results by using an efficient method that extracts the candidate targets in the pre-processing stage and fuses the local scale, blob-based contrast map and gradient map in the detection stage. We also developed mid-wave infrared (MWIR) and long-wave infrared (LWIR) cameras for data collection experiments and algorithm evaluations. Experimental results using both publicly available datasets and image sequences acquired by our cameras clearly demonstrated that the proposed method achieves high detection accuracy with the mean AUC being at least 22.3% higher than comparable methods, and the computational cost beating the other methods by a large margin. © 2021 by the authors. Licensee MDPI, Basel, Switzerland.Accession Number: 20214711200384 -
Record 44 of
Title:Spectral-temporal channeled spectropolarimetry using deep-learning-based adaptive filtering
Author(s):Li, Qiwei(1,2,3); Song, Jiawei(2); Alenin, Andrey S.(2); Tyo, J. Scott(2)Source: Optics Letters Volume: 46 Issue: 17 DOI: 10.1364/OL.436031 Published: September 1, 2021Abstract:Channeled spectropolarimetry (CSP) employing low-pass channel extraction filters suffers from cross talk and spectral resolution loss. These are aggravated by empirically defining the shape and scope of the filters for different measured. Here, we propose a convolutional deep-neural-networkbased channel filtering framework for spectrally-temporally modulated CSP. The network is trained to adaptively predict spectral magnitude filters (SMFs) that possess wide bandwidths and anti-cross-talk features that adapt to scene data in the two-dimensional Fourier domain. Mixed filters that combine the advantages of low-pass filters and SMFs demonstrate superior performance in reconstruction accuracy. ©2021 Optical Society of America.Accession Number: 20213610869064 -
Record 45 of
Title:Simple reformulation of the coordinate transformation method for gratings with a vertical facet or overhanging profile
Author(s):Ming, Xianshun(1,2); Sun, Liqun(2)Source: Applied Optics Volume: 60 Issue: 15 DOI: 10.1364/AO.423209 Published: May 20, 2021Abstract:We reformulate the coordinate transformation method (C method) for gratings with a vertical facet or overhanging profile (overhanging gratings), in which no tensor concept is involved, only the knowledge of elementary mathematics and Maxwell’s equations in the rectangular coordinate system is used, and we provide a detailed recipe for programming. This formulation is easy to understand and implement. It adopts the strategy of a rotating coordinate system from Plumey et al. [J. Opt. Soc. Am. A 14, 610 (1997)] and expresses it with the method of changing variables from Li et al. [Appl. Opt. 38, 304 (1999)]. We investigate several typical overhanging gratings by the reformulated C method, and we validate and compare the results with the Fourier modal method, which shows that it is superior, especially for metal deep smooth gratings. This reformulation can facilitate the research in light couplers for optical engineers. © 2021 Optical Society of AmericaAccession Number: 20212110396436 -
Record 46 of
Title:Development of point diffraction interferometer by a dimension-reduction-based phase-shifting algorithm
Author(s):Feng, Leijie(1); Du, Hubing(1); Liu, Chang(1); Han, Jinlu(1); Zhang, Gaopeng(2); Wang, Feng(2); Zhao, Zixin(3); Gao, Fen(1)Source: Applied Optics Volume: 60 Issue: 30 DOI: 10.1364/AO.439512 Published: October 20, 2021Abstract:To avoid exhaustive calibration of the shifter device in point diffraction interferometers, we present a dimension-reduction-based method to reconstruct the phase map from more phase-shifting fringe patterns with three or more frames. The proposed method assumes that the intensity space can be described adequately by the sine and cosine of multiple phase shifts introduced, which are the basis of the intensity space. Then, low-dimensional approximations of high-dimensional intensity spaces are determined by the newly developed reduced basis decomposition technique. Finally, the phase is reconstructed using the low-dimensional surrogates of the intensity spaces without the knowledge of accurate phase steps. Numerical and experimental studies demonstrated that the proposed method outperforms the existing popular phase reconstruction techniques in terms of accuracy and efficiency. Moreover, the performance of the proposed method is not limited by variations in the background and modulation, unlike the existing phase-shifting-algorithm-based approaches. © 2021 Optical Society of AmericaAccession Number: 20214411085315 -
Record 47 of
Title:Resolution-enhanced phase retrieval for fringe reflection technology with structured light illumination
Author(s):Shen, Xianmeng(1,2); Ren, Maoyun(1,2); Ma, Suodong(1,2,3); Xu, Shengzhi(1,2); Liu, Mingrui(1,2); Zheng, Lichen(1,2)Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 11899 Issue: DOI: 10.1117/12.2603764 Published: 2021Abstract:Owing to its advantages of simple system structure, large dynamic range and high measurement accuracy, fringe reflection technique (FRT) is becoming a powerful tool for specular free-form surface testing in the fields of reverse engineering, defect inspection, optical manufacturing, etc. However, due to the optical transfer function (OTF) of the FRT optical system, high-frequency information on the surface of an element under test is easily lost, which affects the high-precision acquisition of three-dimensional (3D) topography especially in microscopic measurement. Although the above problem can be suppressed to some extent by using or designing high-performance optical systems, the significant increase in the complexity and cost of the measurement system is sometimes unacceptable. To overcome the afore mentioned issue, a resolution-enhanced phase retrieval algorithm based on structured illumination microscopy (SIM) for FRT with an ordinary optical system is proposed in this paper. The combination of FRT and SIM is realized by projecting conventional phase-shifting fringe patterns in multiple directions. In principle, resolution-enhanced phase retrieval with super-diffraction limitation (up to twice the pass band of OTF) can be realized through spectrum extraction, stitching and shifting. A low cost, compact, coaxial FRT setup based on open-source hardware is designed and built for experimental verification. Simulations and experimental results demonstrate the effectiveness of the proposed technique. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.Accession Number: 20215211394376 -
Record 48 of
Title:Nonseparable modulation strategy for channeled spatiotemporal Stokes polarimeters
Author(s):Li, Qiwei(1,2); Song, Jiawei(3); Alenin, Andrey S.(3); Scott Tyo, J.(3)Source: Applied Optics Volume: 60 Issue: 3 DOI: 10.1364/AO.412963 Published: January 20, 2021Abstract:Spatiotemporally modulated polarimeters have shown promising imaging performance by leveraging the tradeoff between spatial bandwidth and temporal bandwidth to outperform polarimeters that use spatial or temporal modulation alone. However, the existing separable modulation strategy, in which the spatial carriers are generated independently from the temporal carriers, makes such devices sensitive to the systematic errors of the rotation element inevitably. In this paper, we propose two novel strategies that have spatiotemporal modulation that is inherently mixed. The method enables different elements of the Mueller matrix to be used to create the carriers, reducing the effects of systematic errors in different ways. We present the indepth comparison of the channel structure and the reconstruction accuracy of each modulation strategy in various bandwidth scenarios under the presence of systematic error. Simulation results show that the nonseparable modulation can provide higher reconstruction accuracy of polarimetric information as compared to the separable strategy. © 2021 Optical Society of AmericaAccession Number: 20211310144772