2013

2013

  • Record 373 of

    Title:Shock response of a rubber double-deck vibration isolation system
    Author(s):Zhang, Zhi(1); Zhang, Zhao-Hui(1); Dong, Sen(1); Wang, Ze-Feng(1); Wu, Li(1); Zhang, Kai-Sheng(1); Cao, Jian-Zhong(1)
    Source: Zhendong yu Chongji/Journal of Vibration and Shock  Volume: 32  Issue: 23  DOI:   Published: 2013  
    Abstract:In aerospace, mechanical environment is abominable. Aiming at this extreme environment, shock response of a double-deck isolation system was studied, and its shock response spectrum under 8000g acceleration shock was analyzed. A rubber damper was used to improve the shock resistance ability of the system, and several factors affecting the shock resistance ability of the rubber damper were analyzed. The test results showed that under 8000g acceleration shock, the shock resistance ability of the double-deck isolation system is better than that of a single-deck isolation system; meanwhile, the span of the rubber damper can greatly affect the shock resistance ability of the system. The study results were very important for shock resistance analysis and structural design of aerospace products.
    Accession Number: 20140417231281
  • Record 374 of

    Title:Three-dimensional structure trajectory guiding scan and points cloud registration
    Author(s):Cai, Yong(1)
    Source: Applied Mechanics and Materials  Volume: 300-301  Issue:   DOI: 10.4028/www.scientific.net/AMM.300-301.423  Published: 2013  
    Abstract:Three-dimensional visual scanning is an advanced non-contact measurement approach that can obtain a complete surface model of object. But, in the method, the local raster point cloud sets calculated must be registered to a whole. To reduce fluctuating errors and increase efficiency, we propose a new method that scanner can be moved around object guided by pre-planning trajectory of the robot. First, considering the structural parameters of the robot, the Inverse Kinematics Problem of each joint rotating function is derived, and the moving trajectory is simulated. Then, the relationship between the distances of controlled motion and data gotten by scanner is analyzed, a transform matrix which registered the local point cloud sets is deduced, the raster data can be normalized to the world coordinate by it. The experimental results show that error of registration is less than 0.09mm. The method is suited to measuring different targets in robot workspace. It can improve the efficiency and flexibility of visual measurement system. © (2013) Trans Tech Publications, Switzerland.
    Accession Number: 20131116115038
  • Record 375 of

    Title:Spectral segmentation via midlevel cues integrating geodesic and intensity
    Author(s):Lu, Huchuan(1); Zhang, Ruixuan(1); Li, Shifeng(1); Li, Xuelong(2)
    Source: IEEE Transactions on Cybernetics  Volume: 43  Issue: 6  DOI: 10.1109/TCYB.2013.2243432  Published: December 2013  
    Abstract:Image segmentation still remains as a challenge in image processing and pattern recognition when involving complex natural scenes. In this paper, we present a new affinity model for spectral segmentation based on midlevel cues. In contrast to most existing methods that operate directly on low-level cues, we first oversegment the image into superpixel images and then integrate the geodesic line edge and intensity cue to form the similarity matrix $W$ so that it more accurately describes the similarity between data. The geodesic line edge could avoid strong boundary and represent the true boundary between two superpixels while the mean red green blue vector could describe the intensity of superpixels better. As far as we know, this is a totally new kind of affinity model to represent superpixels. Based on this model, we use the spectral clustering in the superpixel level and then achieve the image segmentation in the pixel level. The experimental results show that the proposed method performs steadily and well on various natural images. The evaluation comparisons also prove that our method achieves comparable accuracy and significantly performs better than most state-of-the-art algorithms. © 2013 IEEE.
    Accession Number: 20135117094999
  • Record 376 of

    Title:Effects of different ligands on luminescence properties of LaF3: Nd nanoparticles
    Author(s):Gui, Luo(1); Cui, Xiaoxia(2); Wei, Wei(2); Peng, Bo(2); Fan, Dianyuan(1)
    Source: Journal of Rare Earths  Volume: 31  Issue: 7  DOI: 10.1016/S1002-0721(12)60335-4  Published: July 2013  
    Abstract:A series of neodymium-doped lanthanum fluoride nanoparticles (NPs) were synthesized with hydrothermal method, and the effects of several ligands on the luminescence properties of the NPs were investigated. The X-ray diffraction (XRD) results indicated that the crystal phases of the modified NPs coincided with the standard spectrum. The transmission electron microscopy (TEM) showed that the samples were of similar size, shape and dispersibility. The infrared spectra suggested that the content of -OH groups as quenchers on the NPs surfaces decreased after modification. Compared with NPs modified by branched paraffin ligands, NPs conjugating ring-contained modifiers had less quenching effect and possessed stronger fluorescence intensity and longer fluorescence lifetime. © 2013 The Chinese Society of Rare Earths.
    Accession Number: 20133416641526
  • Record 377 of

    Title:A novel color constancy algorithm based on gray-curve regulation
    Author(s):Guo, Huinan(1); Zhou, Zuofeng(1); Tang, Yao(1); Cao, Jianzhong(1)
    Source: Proceedings of SPIE - The International Society for Optical Engineering  Volume: 8878  Issue:   DOI: 10.1117/12.2030932  Published: 2013  
    Abstract:Color constancy performs an important role in computer vision and digital color image processing. The traditional color regulation algorithms which are based on the gray world assumption neglect the correlation between the three stimulus color components and the nonlinear effect between the different gray levels. In this paper, we propose a novel color constancy algorithm which is based on gray-curve regulation. The proposed algorithm first divides the luminance interval into several parts and then regulates the gray-curves of RGB three channels in each subinterval which can solve the correlation and nonlinear problems simultaneously. The experiment results show the effectiveness of the proposed algorithm. © 2013 SPIE.
    Accession Number: 20135017087314
  • Record 378 of

    Title:CFRC outer baffle structure & modal analysis and test of space camera
    Author(s):Chu, Changbo(1); Chai, Wenyi(1); Zhang, Haosu(1); He, Tianbing(1); Wen, Desheng(1)
    Source: Hongwai yu Jiguang Gongcheng/Infrared and Laser Engineering  Volume: 42  Issue: 4  DOI:   Published: April 2013  
    Abstract:Outer baffle of space camera was designed and manufactured with more complicated shape and multilevel inclined vanes using high strengthened carbon fiber reinforced composite (CFRC) T300. Meanwhile the manufacturing process was introduced briefly. Modal analysis was finished with finite element method (FEM). Fixing the outer baffle on an existing fixture the modal test was finished and the radial modes and shapes were obtained, the test frequency response function map was also presented. Comparing the finite element analysis (FEA) and testing data it is indicated that the error of every mode is less than 2.3% between the FEA model & the fact product. It can be seen that the structure of CFRC outer baffle is reasonable and the FEM is feasible. Used into the kinetic modeling process of the space camera, the FEM could have guidance significance to the design of the whole structure of camera.
    Accession Number: 20132616448420
  • Record 379 of

    Title:An L-band graphene-oxide mode-locked fiber laser delivering bright and dark pulses
    Author(s):Zhao, J.Q.(1); Wang, Y.G.(2); Yan, P.G.(1); Ruan, S.C.(1); Zhang, G.L.(1); Li, H.Q.(1); Tsang, Y.H.(3)
    Source: Laser Physics  Volume: 23  Issue: 7  DOI: 10.1088/1054-660X/23/7/075105  Published: July 2013  
    Abstract:In this work we have demonstrated the generation of L-band bright and dark pulses from a graphene-oxide (GO) mode-locked erbium-doped fiber laser. The polymer-hosted graphene-oxide saturable absorber (SA) was fabricated by a vertical evaporation method, which permitted a flexible choice of GO concentration and SA thickness. Stable fundamental solitons with a width of ~770 fs and repetition rate of ~21.79 MHz could be directly generated from the laser cavity. The corresponding spectrum was centered at ~ 1596.0 nm with a 3 dB bandwidth of ~4.454 nm. Triple solitons and fundamentally repetitive dark pulses could also be achieved. By changing the laser cavity design, we also achieved the generation of dark pulses, which had one to three times the repetition rate. The multiple soliton and dark pulse generation were experimentally observed for the first time in such a GO mode-locked fiber laser system, to the best of our knowledge. Our experimental results indicate that the GOSA can be successfully utilized to realize different mode-locked states in the L-band. © 2013 Astro Ltd.
    Accession Number: 20132516432282
  • Record 380 of

    Title:Transfer learning for pedestrian detection
    Author(s):Cao, Xianbin(1); Wang, Zhong(2); Yan, Pingkun(3); Li, Xuelong(3)
    Source: Neurocomputing  Volume: 100  Issue:   DOI: 10.1016/j.neucom.2011.12.043  Published: January 16, 2013  
    Abstract:Most of the existing methods for pedestrian detection work well, only when the following assumption is satisfied: the features extracted from the training dataset and the testing dataset have very similar distributions in the feature space. However, in practice, this assumption does not hold because of the scene complexity and variation. In this paper, a new method is proposed for detecting pedestrians in various scenes based on the transfer learning technique. Our proposed method employs the following two strategies for improving the pedestrian detection performance. First, a new sample screening method based on manifold learning is proposed. The basic idea is to choose samples from the training set, which may be similar to the samples from the unseen scene, and then merge the selected samples into the unseen set. Second, a new classification model based on transfer learning is proposed. The advantage of the classification model is that only a small number of samples need to be used from the unseen scenes. Most of the training samples are still obtained from the training scene, which take up to 90% of the entire training samples. Compared to the traditional pedestrian detection methods, the proposed algorithm can adapt to different scenes for detecting pedestrians. Experiments on two pedestrian detection benchmark datasets, DC and NICTA, showed that the method can obtain better performance as compared to other previous methods. © 2012 Elsevier B.V.
    Accession Number: 20124615668441
  • Record 381 of

    Title:A FPGA-based digital-mirror driving control system for compressive imaging
    Author(s):Li, Shi Bo(1); He, Jun Hua(1); Lv, Pei(1)
    Source: Advanced Materials Research  Volume: 760-762  Issue:   DOI: 10.4028/www.scientific.net/AMR.760-762.1259  Published: 2013  
    Abstract:With FPGA as the host controller, a set of hardware system based on digital micro-mirror device (DMD) for compressive imaging (CI) was developed. This paper studies the principle and characteristic of DMD control. First, the data of PC Screen image is transferred to DMD driver by software using binary pulse width modulation (BPWM) technology. The data is then rewritten into a control single which finally directs the DMD. The test results proved that the digital image received can be loaded onto the micro-mirrors excellently and with optics module installed, the system could successfully implement the CS measurement patterns which satisfy CI system. © (2013) Trans Tech Publications, Switzerland.
    Accession Number: 20134416937961
  • Record 382 of

    Title:Complementary projection hashing
    Author(s):Jin, Zhongming(1); Hu, Yao(1); Lin, Yue(1); Zhang, Debing(1); Lin, Shiding(2); Cai, Deng(1); Li, Xuelong(3)
    Source: Proceedings of the IEEE International Conference on Computer Vision  Volume:   Issue:   DOI: 10.1109/ICCV.2013.39  Published: 2013  
    Abstract:Recently, hashing techniques have been widely applied to solve the approximate nearest neighbors search problem in many vision applications. Generally, these hashing approaches generate 2c buckets, where c is the length of the hash code. A good hashing method should satisfy the following two requirements: 1) mapping the nearby data points into the same bucket or nearby (measured by the Hamming distance) buckets. 2) all the data points are evenly distributed among all the buckets. In this paper, we propose a novel algorithm named Complementary Projection Hashing (CPH) to find the optimal hashing functions which explicitly considers the above two requirements. Specifically, CPH aims at sequentially finding a series of hyper planes (hashing functions) which cross the sparse region of the data. At the same time, the data points are evenly distributed in the hyper cubes generated by these hyper planes. The experiments comparing with the state-of-the-art hashing methods demonstrate the effectiveness of the proposed method. © 2013 IEEE.
    Accession Number: 20141717631768
  • Record 383 of

    Title:Investigation and control of ultrafast laser-induced isotropic and anisotropic nanoscale-modulated index patterns in bulk fused silica
    Author(s):Stoian, R.(1); Mishchik, K.(1); Cheng, G.(2); Mauclair, C.(1); D'Amico, C.(1); Colombier, J.P.(1); Zamfirescu, M.(3)
    Source: Optical Materials Express  Volume: 3  Issue: 10  DOI: 10.1364/OME.3.001755  Published: 2013  
    Abstract:Ultrafast laser-induced refractive index changes in a-SiO2 consist, depending on the irradiation conditions, of either positive variations, voids, or regular nanoscale patterns, each of these underlying specific structural transformations. These allow for obtaining a large palette of optical functions ranging from low loss guiding to anisotropic scattering. While briefly reviewing the excitation mechanisms, we spectroscopically interrogate local electronic and structural transformations of the glass in the isotropic index zones and in the regular self-organized nanostructures, indicating bond breaking and matrix oxygen deficiency. A spatial defect segregation marks the material transformation in the different photoinscription regimes. We equally propose a method of real time control of nanogratings formation under the action of ultrashort laser pulse with variable envelopes. Application as polarizing optical devices is discussed. © 2013 Optical Society of America.
    Accession Number: 20134817027115
  • Record 384 of

    Title:Development of streak camera with anisotropic focusing electron optical system
    Author(s):Tian, J.(1); Ding, Y.(2); Cao, X.(1); Liu, S.(2); Xu, X.(1); Hu, X.(2); Wen, W.(1); Wang, J.(1); Wang, C.(1); Liu, H.(1); Dong, G.(1); Zhang, T.(1); Lu, Y.(1); Wang, X.(1); Liu, J.(1,3)
    Source: Proceedings of SPIE - The International Society for Optical Engineering  Volume: 8796  Issue:   DOI: 10.1117/12.2008988  Published: 2013  
    Abstract:In this paper, the anisotropic focusing technique is used to make a novel streak tube. The salient features are the introduction of both temporally focusing electrodes and spatially focusing electric quadrupole lens. The simulation showed that physical temporal dispersion of 0.38 ps and edge spatial resolution of 56 lp/mm can be achieved. The Nd:YLF 8ps pulse laser was used to calibrate the performance index of streak camera. The static and dynamic spatial resolutions are 35 lp/mm and 25 lp/mm respectively. The dynamic range more than 950:1 and time resolution 8ps can be reached. Furthermore, the magnifications in slit and scanning direction can be adjusted respectively, so it is very convenient to select amplification needed when it is coupled with KB microscope. © 2013 SPIE.
    Accession Number: 20133216576151