-
Record 289 of
Title:Unsupervised feature learning for scene classification of high resolution remote sensing image
Author(s):Fu, Min(1,2); Yuan, Yuan(1); Lu, Xiaoqiang(1)
Source: 2015 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Proceedings Volume: Issue: DOI: 10.1109/ChinaSIP.2015.7230392 Published: August 31, 2015
Abstract:Due to the rapid development of various satellite sensors, a large amount of high resolution remote sensing images can be obtained. In order to efficiently represent the scenes from these high resolution images, an unsupervised feature learning method is proposed for high resolution image scene classification. In the proposed method, a set of filter banks are learned in an unsupervised manner from the unlabeled image patches, which are robust, efficient and do not need elaborately designed descriptors such as SIFT. And then, each image is encoded by these filter banks using a soft distance assignment scheme, generating a final feature vector to excellently represent the image scene. Finally, by virtue of the traditional SVM classifier, the sematic concepts of different scenes can be categorized. Experimental evaluation on the the high resolution remote sensing images demonstrates the effectiveness and good performance of the proposed method. ? 2015 IEEE.
Accession Number: 20160701912158
-
Record 290 of
Title:How to represent scenes for classification?
Author(s):Shi, Jianhua(1,2); Li, Xuelong(1); Dong, Yongsheng(1)
Source: 2015 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Proceedings Volume: Issue: DOI: 10.1109/ChinaSIP.2015.7230389 Published: August 31, 2015
Abstract:Object-based scene image representations can effectively capture the semantic meanings of a scene. However, they usually neglect a scene's structure information. In this paper, we propose a novel and effective detector-based scene representation method for scene classification. In particular, we extract object features by object detectors. By sensible principal component analysis, we obtain a compact representation vector of objects in a scene image. To capture the scene layout, we then train lots of deformable part models to form a scene response vector. By concatenating these two vectors we use a linear support vector machine for scene classification. When combining with DeCAF [1] in a special way, our method is even more powerful on complex scene categorization. Experimental results on the MIT indoor database show that our approach achieves state-of-The-Art performance on scene classification compared with several popular methods. ? 2015 IEEE.
Accession Number: 20160701912155
-
Record 291 of
Title:Design of infrared diffractive telescope imaging optical systems
Author(s):Zhang, Zhoufeng(1,2,3); Hu, Bingliang(1); Yin, Qinye(2); Xie, Yongjun(1); Kang, Fuzeng(1); Wang, Yanjun(1)
Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 9676 Issue: DOI: 10.1117/12.2201125 Published: 2015
Abstract:Diffractive telescope is an updated imaging technology, it differs from conventional refractive and reflective imaging system, which is based on the principle of diffraction image. It has great potential for developing the larger aperture and lightweight telescope. However, one of the great challenges of design this optical system is that the diffractive optical element focuses on different wavelengths of light at different point in space, thereby distorting the color characteristics of image. In this paper, we designs a long-wavelength infrared diffractive telescope imaging system with flat surface Fresnel lens and cancels the infrared optical system chromatic aberration by another flat surface Fresnel lens, achieving broadband light(from 8μm-12μm) to a common focus with 4.6° field of view. At last, the diffuse spot size and MTF function provide diffractive-limited performance. ? 2015 SPIE.
Accession Number: 20155201717484
-
Record 292 of
Title:On two heuristic viewpoints concerning the study of light
Author(s):Bi, Siwen(1,2)
Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 9449 Issue: DOI: 10.1117/12.2085041 Published: 2015
Abstract:It has been a debatable problem that what the essence of light is, and how it is produced. Since the modern times, as James Clerk Maxwell setting the theory of electromagnetic up, the mainstream consciousness was occupied gradually by the wave theory of light. But at the end of 19th century, a series experimental phenomenon weren't precisely explained by the wave theory of light such as photoelectric effect experiment. Then Albert Einstein published his famous paper On a Heuristic Viewpoint Concerning the Production and Transformation of Light, which laid the foundation of light quantum hypothesis. While solving these problems perfectly, a new problem was caused that because the wave theory and the quantum theory are both applicable to interpret some of the experiment of light, what is the essence of light. This paper first outlines the history of optical development and current status, and states the difficulties and deficiencies of the study of light. Then we put forward the key concept of the paper called lightstring which consults some points of the theory of modern optics and physics which called the optical frequency comb and the string theory, then presents the essence of light based on the light string concept in order to make the concept of photons specific. And then we put forward the production mechanism of light - - the String-Light effect based on the concept of light string. In this paper, we attempt to put forward a new idea of the study of the essence of light and the production mechanism of it. ? 2015 SPIE.
Accession Number: 20151100635534
-
Record 293 of
Title:A Variational Approach to Simultaneous Image Segmentation and Bias Correction
Author(s):Zhang, Kaihua(1); Liu, Qingshan(1); Song, Huihui(1); Li, Xuelong(2)
Source: IEEE Transactions on Cybernetics Volume: 45 Issue: 8 DOI: 10.1109/TCYB.2014.2352343 Published: August 1, 2015
Abstract:This paper presents a novel variational approach for simultaneous estimation of bias field and segmentation of images with intensity inhomogeneity. We model intensity of inhomogeneous objects to be Gaussian distributed with different means and variances, and then introduce a sliding window to map the original image intensity onto another domain, where the intensity distribution of each object is still Gaussian but can be better separated. The means of the Gaussian distributions in the transformed domain can be adaptively estimated by multiplying the bias field with a piecewise constant signal within the sliding window. A maximum likelihood energy functional is then defined on each local region, which combines the bias field, the membership function of the object region, and the constant approximating the true signal from its corresponding object. The energy functional is then extended to the whole image domain by the Bayesian learning approach. An efficient iterative algorithm is proposed for energy minimization, via which the image segmentation and bias field correction are simultaneously achieved. Furthermore, the smoothness of the obtained optimal bias field is ensured by the normalized convolutions without extra cost. Experiments on real images demonstrated the superiority of the proposed algorithm to other state-of-the-art representative methods. ? 2013 IEEE.
Accession Number: 20153001072170
-
Record 294 of
Title:Nonlinear optical absorption tuning in Bi3.15Nd0.85Ti3O12 ferroelectric thin films by thickness
Author(s):Li, S.(1,2); Zhong, X.L.(1,2); Cheng, G.H.(3); Liu, X.(3); Zhang, Y.(1,2); Wang, J.B.(1,2); Song, H.J.(1,2); Tan, C.B.(1,2); Li, B.(1,2)
Source: Applied Physics Letters Volume: 106 Issue: 14 DOI: 10.1063/1.4917223 Published: April 6, 2015
Abstract:The tunability of nonlinear optical (NLO) absorption in Bi3.15Nd0.85Ti3O12 (BNT) ferroelectric thin films was investigated through the open aperture Z-scan method with femtosecond laser pulses at the wavelength of 800 nm. NLO absorption responses of the BNT films were observed to be highly sensitive to the film thickness. As the film thickness increases from 106.8 to 139.7 nm, the NLO absorption changes from saturable absorption (SA) to reverse saturable absorption (RSA). When the film thickness further increases to 312.9 nm, the RSA effect is enhanced. A band-gap-related competition between the ground-state excitation and the two-photon absorption is responsible for the absorption switching behavior. Such a tunable NLO absorption can widen the photonic application of the BNT thin films. ? 2015 AIP Publishing LLC.
Accession Number: 20154301422554
-
Record 295 of
Title:Design of a Solid-Core Large-Mode-Area Bragg Fiber
Author(s):Li, Lu(1,4); Pang, Li-Hui(2); Zhou, Zhi-Guang(1); Zhang, Ai-Dong(1); He, Jian-Li(1); Si, Jin-Hai(4); Lin, Ao-Xiang(3)
Source: Chinese Physics Letters Volume: 32 Issue: 5 DOI: 10.1088/0256-307X/32/5/054201 Published: May 1, 2015
Abstract:We report the design and simulation results of a solid-core Bragg fiber with 3-bilayer periodic cladding. The simulation results present single mode bandgap guidance, a large effective area of ～400 μm2 around 1.08 μm, a very low bend loss of 0.038 dB/m at 1.08 μm even under tight bend radius (R = 4 cm), and excellent beam quality. The results indicate that the proposed fiber could be a competitive solution for high-power fiber laser applications. ? 2015 Chinese Physical Society and IOP Publishing Ltd.
Accession Number: 20220711658715
-
Record 296 of
Title:Low-loss slow-light in periodic plasmonic waveguides
Author(s):Zhang, Lingxuan(1,2); Lu, Xiaoyuan(1,2); Gong, Yongkang(1,3); Copner, Nigel(1,3); Zhao, Wei(1,2); Wang, Guoxi(1,2); Zhang, Wenfu(1,2)
Source: IEEE Photonics Technology Letters Volume: 27 Issue: 11 DOI: 10.1109/LPT.2015.2414939 Published: June 1, 2015
Abstract:We have proposed and investigated two kinds of waveguide systems (i.e., dielectric-dielectric-metal and dielectric-metal-dielectric waveguides) which are able to support low-loss slow-light guided mode at telecommunication regime of 1.55 μm. Simulation results demonstrate that the propagation length of the proposed waveguides is as long as tens or even hundreds of micrometers, which is more than ten times higher than that of the traditional metal-dielectric-metal waveguides. Moreover, the proposed waveguides have better slow light performance and a figure of merit that can reach a value as high as 18 000. In addition, the slow-down factor can be flexibly tuned by adjusting the geometrical parameters. The proposed waveguide have advantages of low loss and compact configuration, which can find potential applications, such as optical buffers, in highly integrated optical communication systems. ? 2015 IEEE.
Accession Number: 20152100859707
-
Record 297 of
Title:Performances of a solid streak camera based on conventional CCD with nanosecond time resolution
Author(s):Wang, Bo(1); Bai, Yonglin(1); Zhu, Bingli(1); Gou, Yongsheng(1); Xu, Peng(2); Bai, Xiaohong(3); Liu, Baiyu(1); Qin, Junjun(1)
Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 9449 Issue: DOI: 10.1117/12.2085044 Published: 2015
Abstract:Imaging systems with high temporal resolution are needed to study rapid physical phenomena ranging from shock waves, including extracorporeal shock waves used for surgery, to diagnostics of laser fusion and fuel injection in internal combustion engines. However, conventional streak cameras use a vacuum tube making thus fragile, cumbersome and expensive. Here we report an CMOS streak camera project consists in reproducing completely this streak camera functionality with a single CMOS chip. By changing the mode of charge transfer of CMOS image sensor, fast photoelectric diagnostics of single point with linear CMOS and high-speed line scanning with array CMOS sensor can be achieved respectively. A fast photoelectric diagnostics system has been designed and fabricated to investigate the feasibility of this method. Finally, the dynamic operation of the sensors is exposed. Measurements show a sample time of 500 ps and a time resolution better than 2 ns.. ? 2015 SPIE.
Accession Number: 20151100635608
-
Record 298 of
Title:Scanning stereomicroscopy with two-photon excitation and scanned Bessel beams
Author(s):Yang, Yanlong(1); Zhou, Xing(1); Li, Runze(1); Peng, Tong(1); Lei, Ming(1); Wu, Di(1); Yao, Baoli(1); Van Horn, Mark(2); Chen, Xun(2); Ye, Tong(2)
Source: 2015 Opto-Electronics and Communications Conference, OECC 2015 Volume: Issue: DOI: 10.1109/OECC.2015.7340281 Published: November 30, 2015
Abstract:Bessel beams have been used in many applications due to their unique optical properties of maintaining their intensity profiles unchanged during propagation. In imaging applications, Bessel beams have been successfully used to provide extended focuses for volumetric imaging and uniformed illumination plane in light-sheet microscopy. Coupled with two-photon excitation, Bessel beams have been successfully used in realizing fluorescence projected volumetric imaging. We demonstrated previously a stereoscopic solution-two-photon fluorescence stereomicroscopy (TPFSM)-for recovering the depth information in volumetric imaging with Bessel beams. In TPFSM, tilted Bessel beams were used to generate stereoscopic images on a laser scanning two-photon fluorescence microscope; upon post image processing we could successfully provide 3D perception of acquired volume images by wearing anaglyph 3D glasses. However, tilted Bessel beams were generated by shifting either an axicon or an objective laterally; the slow imaging speed and severe aberrations made it hard to use in real-time volume imaging. In this article, we report recent improvements of TPFSM with newly designed scanner and imaging software, which allows 3D stereoscopic imaging without moving any of the optical components on the setup. This improvement has dramatically improved focusing qualities and imaging speed so that the TPFSM can be performed potentially in real-time to provide 3D visualization in scattering media without post image processing. ? 2015 IEEE.
Accession Number: 20161202125478
-
Record 299 of
Title:Superpixel fats for fast foreground extraction
Author(s):Liu, Kang(1,2); Li, Xuelong(1); Dong, Yongsheng(1)
Source: 2015 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2015 - Proceedings Volume: Issue: DOI: 10.1109/ChinaSIP.2015.7230377 Published: August 31, 2015
Abstract:Fast foreground extraction is an important and challenging problem. Although GrabCut can perform well in foreground extraction, the average accuracy is not satisfactory, and more importantly, its computational cost is large. In this paper, we propose a fast interactive foreground extraction based on superpixel GrabCut and matting. Specifically, we use a superpixel method to process images in order to improve the efficiency of the GrabCut. To obtain a refined mask, we employ the Shared matting method whose trimaps are implemented by our constructed fast and adaptive trimaps (FATs). Experimental results demonstrate the proposed method is more competitive and effective in the visual sense, E-Time and MSE criterions. ? 2015 IEEE.
Accession Number: 20160701912143
-
Record 300 of
Title:The design and implementation for a range-gated ICCD
Author(s):He, Huan(1,2); Tian, Jin-Shou(1); Wen, Wen-Long(1); Wang, Jun-Feng(1); Liu, Hu-Lin(1); Sai, Xiao-Feng(1); Wei, Yong-Lin(1); Xin, Li-Wei(1); Li, Li-Zhou(3)
Source: Guangzi Xuebao/Acta Photonica Sinica Volume: 44 Issue: 6 DOI: 10.3788/gzxb20154406.0611001 Published: June 1, 2015
Abstract:A high-speed gated second generation image intensifier which is of short afterglow, high resolution and fast response time was designed. A high performance range-gated ICCD was obtained by coupling a fiber taper to CCD. The performance of each component and their effects on the system spatial resolution were analyzed. FPGA was employed to design the circuit control system which can achieve digital control for ICCD by producing a nanosecond gating width. The system can also obtain clear images of objectives in different brightness and distance by adjusting the pulse width and delay time. In addition, the background noise was reduced and the dynamic range of imaging was increased. The gain of this system can be monitored and adjusted to reach a signal-noise ratio of 20∶1 dB. A gating DC pulse with voltage of -200 V and width of 3 ns-DC continuously adjustable was applied to achieve image intensifier gating. In order to improve the supplement speed of electrons to the photocathode, a square grid which has 5 μm line width and 50 μm spacing lithography was equipped to the input window surface. This also ensured that the photocathode response speed is fast enough when the gating width is 3 ns. The highest gating frequency can reach to 300 kHz. The experimental test shows that the gain of the image intensifier is 10 718 cd/m2 lx when Microchannel Plate voltage is 700 V and phosphor screen voltage is 5 000 V, the system spatial resolution is 29.7 lp/mm. ?, 2015, Chinese Optical Society. All right reserved.
Accession Number: 20152901043464
