2014
2014
-
Record 337 of
Title:A new sparse feature-based patch for dense correspondence
Author(s):Qing, Xiameng(1); Shen, Jianbing(1); Li, Xuelong(2); Jia, Yunde(1)Source: Proceedings - IEEE International Conference on Multimedia and Expo Volume: 2014-September Issue: Septmber DOI: 10.1109/ICME.2014.6890288 Published: September 3, 2014Abstract:This paper presents a new method to compute the dense correspondences between two images by using the sparse feature-based patches in an energy optimization framework. Many transformation and deformation cues such as color, scale and rotation should be considered when we finding dense correspondences between images. However, most existing methods only consider part of these transformations, which will introduce the uncorrect correspondence results. In terms of the property of the sparse feature and the principle that nearest sub-scenes and neighbors are much more similar, we design a new energy optimization to guide the dense matching process. Both transformation and deformation are considered in our energy optimization framework since we design the feature-based patches. Thus, our algorithm can match the complicated scenes and objects robustly. At last, a local refinement technique is proposed to solve the perturbation of the matched patches. Experimental results demonstrate that our method outperforms the state-of-the-art algorithms. ? 2014 IEEE.Accession Number: 20153001066592 -
Record 338 of
Title:Toward scalable systems for big data analytics: A technology tutorial
Author(s):Hu, Han(1,2); Wen, Yonggang(1,3); Chua, Tat-Seng(1,2); Li, Xuelong(1,4)Source: IEEE Access Volume: 2 Issue: DOI: 10.1109/ACCESS.2014.2332453 Published: 2014Abstract:Recent technological advancements have led to a deluge of data from distinctive domains (e.g., health care and scientific sensors, user-generated data, Internet and financial companies, and supply chain systems) over the past two decades. The term big data was coined to capture the meaning of this emerging trend. In addition to its sheer volume, big data also exhibits other unique characteristics as compared with traditional data. For instance, big data is commonly unstructured and require more real-time analysis. This development calls for new system architectures for data acquisition, transmission, storage, and large-scale data processing mechanisms. In this paper, we present a literature survey and system tutorial for big data analytics platforms, aiming to provide an overall picture for nonexpert readers and instill a do-it-yourself spirit for advanced audiences to customize their own big-data solutions. First, we present the definition of big data and discuss big data challenges. Next, we present a systematic framework to decompose big data systems into four sequential modules, namely data generation, data acquisition, data storage, and data analytics. These four modules form a big data value chain. Following that, we present a detailed survey of numerous approaches and mechanisms from research and industry communities. In addition, we present the prevalent Hadoop framework for addressing big data challenges. Finally, we outline several evaluation benchmarks and potential research directions for big data systems. ? 2014 IEEE.Accession Number: 20144000079079 -
Record 339 of
Title:An Ytterbium-doped fiber laser with dark and Q-switched pulse generation using graphene-oxide as saturable absorber
Author(s):Zhao, Junqing(1); Wang, Yonggang(2); Yan, Peiguang(1); Ruan, Shuangchen(1); Tsang, Yuen(3); Zhang, Gelin(1); Li, Huiquan(1)Source: Optics Communications Volume: 312 Issue: DOI: 10.1016/j.optcom.2013.09.038 Published: February 1, 2014Abstract:We demonstrate an ytterbium-doped fiber (YDF) laser being able to generate dark pulses and Q-switched pulses by using a graphene oxide (GO) based composite as saturable absorber (SA). These operation regimes are realized for the first time in a 1-μm band fiber laser as far as our best knowledge. In the fabrication of GO based saturable absorber (GOSA), a vertical evaporation process is included, which results in a GO concentration and thickness variable SA. The generated dark pulse has pulse duration in ns-scale. In the Q-switched operation, pulses with durations from 5.321 μs to 2.655 μs and repetition rates from 28.47 kHz to 58.95 kHz can be obtained as the launched pump power is changed. The transition from the dark pulse to Q-switched operations can be clearly seen as the pump power is increased to a certain value. The radio frequency (RF) and optical spectrum variations with respect to the change of pump power have also been analyzed in detail. ? 2013 Elsevier B.V. All rights reserved.Accession Number: 20144800244573 -
Record 340 of
Title:Improving level set method for fast auroral oval segmentation
Author(s):Yang, Xi(1); Gao, Xinbo(1); Tao, Dacheng(2); Li, Xuelong(3)Source: IEEE Transactions on Image Processing Volume: 23 Issue: 7 DOI: 10.1109/TIP.2014.2321506 Published: July 2014Abstract:Auroral oval segmentation from ultraviolet imager images is of significance in the field of spatial physics. Compared with various existing image segmentation methods, level set is a promising auroral oval segmentation method with satisfactory precision. However, the traditional level set methods are time consuming, which is not suitable for the processing of large aurora image database. For this purpose, an improving level set method is proposed for fast auroral oval segmentation. The proposed algorithm combines four strategies to solve the four problems leading to the high-time complexity. The first two strategies, including our shape knowledge-based initial evolving curve and neighbor embedded level set formulation, can not only accelerate the segmentation process but also improve the segmentation accuracy. And then, the latter two strategies, including the universal lattice Boltzmann method and sparse field method, can further reduce the time cost with an unlimited time step and narrow band computation. Experimental results illustrate that the proposed algorithm achieves satisfactory performance for auroral oval segmentation within a very short processing time. ? 2014 IEEE.Accession Number: 20142317782048 -
Record 341 of
Title:Flux-solvothermal preparation of dispersible LiLa0.4Nd 0.6(PO3)4 microcrystals with regular morphology and superior fluorescence
Author(s):Wang, Zhongyue(1); Duan, Weikuan(1); Cui, Xiaoxia(3); Liang, Chen(1); Zheng, Ruilin(2); Wei, Wei(2,3)Source: Journal of Materials Chemistry C Volume: 2 Issue: 20 DOI: 10.1039/c3tc32438a Published: May 28, 2014Abstract:A flux-solvothermal method was used to grow a type of novel dispersible LiLa0.4Nd0.6(PO3)4 microcrystals with high Nd3+ ions concentration of 2.63 × 1021 cm-3 and excellent fluorescence properties for the first time. By optimizing experimental conditions, microcrystals with sizes in the region of 1.5-5 μm, a strong emission intensity and long lifetime of 107 μs were obtained. The results show that their transparent dispersion in the mixed solvents of DMSO and CHBr2CHBr2 had a strong absorption at 800 nm, low solvent quenching ratio of 6.5%, high quantum yield of 32.17% and large emission cross section of 4.39 × 10-20 cm2 when the Nd3+ ions concentration is 1 × 1020 cm -3, which imply the microcrystals are of potential application in transparent glass-ceramics, dispersion amplifiers and lasers. ? 2014 the Partner Organisations.Accession Number: 20141917690372 -
Record 342 of
Title:A unified learning framework for single image super-resolution
Author(s):Yu, Jifei(1); Gao, Xinbo(1); Tao, Dacheng(2); Li, Xuelong(3); Zhang, Kaibing(4)Source: IEEE Transactions on Neural Networks and Learning Systems Volume: 25 Issue: 4 DOI: 10.1109/TNNLS.2013.2281313 Published: April 2014Abstract:It has been widely acknowledged that learning- and reconstruction-based super-resolution (SR) methods are effective to generate a high-resolution (HR) image from a single low-resolution (LR) input. However, learning-based methods are prone to introduce unexpected details into resultant HR images. Although reconstruction-based methods do not generate obvious artifacts, they tend to blur fine details and end up with unnatural results. In this paper, we propose a new SR framework that seamlessly integrates learning- and reconstruction-based methods for single image SR to: 1) avoid unexpected artifacts introduced by learning-based SR and 2) restore the missing high-frequency details smoothed by reconstruction-based SR. This integrated framework learns a single dictionary from the LR input instead of from external images to hallucinate details, embeds nonlocal means filter in the reconstruction-based SR to enhance edges and suppress artifacts, and gradually magnifies the LR input to the desired high-quality SR result. We demonstrate both visually and quantitatively that the proposed framework produces better results than previous methods from the literature. ? 2012 IEEE.Accession Number: 20141317522843 -
Record 343 of
Title:Optical system design of Lander camera with focusing type
Author(s):He, Ying-Hong(1); Yang, Jian-Feng(1); Ma, Xiao-Long(1); Lü, Juan(1); Xue, Bin(1); Li, Ting(1); Jiao, Ying(1)Source: Guangzi Xuebao/Acta Photonica Sinica Volume: 43 Issue: 4 DOI: 10.3788/gzxb20144304.0422004 Published: April 2014Abstract:A optical system of lander optical payload was designed for the requirement of focusing on objects. A noval focusing type was made that the optical system could keep fixed focal length during focusing. An approaching telecentric optical structure is designed, the optical system was made up of front fixing lens unit and rear focusing lens unit. The front fixing lens is a kind of afocal optical system. The rear focusing lens is at the back of fixing lens unit. The optical can image objects from 0.8 m to infinity in focus by moving rear focusing lens unit. The optical system with spectral range from 400 to 1 000 nm, fixed focal length of 50 mm and F number of 8 can focus on objects located within the range of 0.8 meter to infinity. At last, The modulation transfer function of the optical system can reach to diffraction limit, and distortion is less 1%. The image quality can meet the requirement to the optical system. By the analysis of optical system tolerance, the optical system can keep good imaging quality after fabrication. The optical system can be used to explore moon, mars and asteroids.Accession Number: 20142117743311 -
Record 344 of
Title:Design of 2-D reflection mirror gimbal for lunar based astronomy optical telescope
Author(s):Huang, Jing(1); Liu, Zhao-Hui(1); Xie, You-Jin(1); Li, Zhi-Guo(1); Wang, Zhen-Yu(1)Source: Guangxue Jingmi Gongcheng/Optics and Precision Engineering Volume: 22 Issue: 8 DOI: 10.3788/OPE.20142208.2173 Published: August 1, 2014Abstract:To implement the astronomical observation on the moon, the thermal analysis and structure design of a 2D reflection mirror gimbal for the lunar based astronomy optical telescope were performed. The 2D rotary structure of reflection mirror was designed based on lightweighting method to reduce the load mass. Because an external rotor mechanism was used for the vertical shaft, the first order mode of the system along the emission direction was greatly improved. For a larger span in the horizontal shaft, a shaft system with one fixed end and another free end was adopted. Moreover, the clearance of deep groove ball bearings was designed rationally to eliminate the block of the rotary structure caused by temperature changes on the moon. In order to meet the high accuracy, the reflection mirror gimbal used a worm gear and a step motor to drive the gimbal and to control the machining in a strict technology. In addition, an optics switch was used to achieve high accuracy position and make the accuracy of gimbal be better than 60″. Experiments show that the first order resonant frequency of the system can reach 81 Hz along the rocket-firing track, the mechanism works well at -25℃ to + 60℃ without stuck phenomenon and its direction accuracy is less than 60″. These results demonstrate that the system is characterized by high accuracy, high reliability and light weight.Accession Number: 20143900075538 -
Record 345 of
Title:A method for ShenguangIII diagnostic instrument manipulator high precision automatic aiming
Author(s):Wang, Wei(1); Wang, Zheng-Zhou(1); He, Jun-Hua(1); Zhang, Min(1); Wei, Ming-Zhi(1); Yan, Ya-Dong(1)Source: Guangzi Xuebao/Acta Photonica Sinica Volume: 43 Issue: 6 DOI: 10.3788/gzxb20144306.0612002 Published: June 2014Abstract:In order to realize diagnostic instrument manipulator precise targeting, an aiming method based on visual servoing was proposed. Defined the target deviation and projection matrix in the vision system, the calculated target deviation value computed by projection matrix was closed to the true value. Using a 3-DOF hybrid mechanism motion mode, the motion accuracy was promoted. With optimizing visual servoing strategy based on projection matrix and motion mode, the total targeting work was greatly simplified, making the vision system calibration unnecessary. Automatic positioning tests indicate that pointing accuracy is 11μm in X direction and 12 μm in Y direction. Radical positioning accuracy is 14 μm. Laser shot tests with streak camera show that the automatic aiming method can satisfy the requirements of project.Accession Number: 20142917960007 -
Record 346 of
Title:Detail enhancement for high dynamic range images rendering based on multi-level decomposition
Author(s):Wang, Hua(1); Zhang, Hui(1); Cao, Jian Zhong(1); Zhou, Zuo Feng(1); Yang, Lei(1); Zhao, Xiao Dong(1); Ren, Long(1)Source: Applied Mechanics and Materials Volume: 543-547 Issue: DOI: 10.4028/www.scientific.net/AMM.543-547.2600 Published: 2014Abstract:Various tone reproduction operators have been proposed to display high dynamic range images on low dynamic range (LDR) devices. Many recent computational photography techniques decompose an image into a piecewise smooth base layer, containing large scale variations in intensity, and a residual detail layer capturing the smaller scale details in the image. In these techniques, it's important to control the scale of the extracted details and it is often needed to manipulate details in order to avoid the appearance of visual artifacts. In this paper, a new method is proposed to preserve details for high dynamic range images tone reproduction using multi-level image decomposition. We show that current base-detail decomposition techniques, based on the bilateral filter, are limited in their ability to extract detail at arbitrary scales. Thus, we achieve detail enhancement by applying the bilateral filter iteratively, and this process is called multi-level decomposition. By minimizing the proposed energy function, we can choose the proper decomposition level. Simulation results demonstrate that the proposed method can acquire better visual quality in detail enhancement while make the base smoothness. ? (2014) Trans Tech Publications, Switzerland.Accession Number: 20141717625176 -
Record 347 of
Title:Robust Hashing with local models for approximate similarity search
Author(s):Song, Jingkuan(1); Yang, Yi(1); Li, Xuelong(2); Huang, Zi(1); Yang, Yang(3)Source: IEEE Transactions on Cybernetics Volume: 44 Issue: 7 DOI: 10.1109/TCYB.2013.2289351 Published: July 2014Abstract:Similarity search plays an important role in many applications involving high-dimensional data. Due to the known dimensionality curse, the performance of most existing indexing structures degrades quickly as the feature dimensionality increases. Hashing methods, such as locality sensitive hashing (LSH) and its variants, have been widely used to achieve fast approximate similarity search by trading search quality for efficiency. However, most existing hashing methods make use of randomized algorithms to generate hash codes without considering the specific structural information in the data. In this paper, we propose a novel hashing method, namely, robust hashing with local models (RHLM), which learns a set of robust hash functions to map the high-dimensional data points into binary hash codes by effectively utilizing local structural information. In RHLM, for each individual data point in the training dataset, a local hashing model is learned and used to predict the hash codes of its neighboring data points. The local models from all the data points are globally aligned so that an optimal hash code can be assigned to each data point. After obtaining the hash codes of all the training data points, we design a robust method by employing 2,1-norm minimization on the loss function to learn effective hash functions, which are then used to map each database point into its hash code. Given a query data point, the search process first maps it into the query hash code by the hash functions and then explores the buckets, which have similar hash codes to the query hash code. Extensive experimental results conducted on real-life datasets show that the proposed RHLM outperforms the state-of-the-art methods in terms of search quality and efficiency. ? 2013 IEEE.Accession Number: 20142717881497 -
Record 348 of
Title:Irregular image adaptive focusing system for diffraction grating
Author(s):Wang, Wei(1); Li, Dong-Jian(1); Li, Hong-Guang(1); Wang, Zheng-Zhou(1); Cao, Shi-Kang(1)Source: Guangxue Jingmi Gongcheng/Optics and Precision Engineering Volume: 22 Issue: 4 DOI: 10.3788/OPE.20142204.0822 Published: April 2014Abstract:The characteristics of faculae for a larger diameter diffraction grating were analyzed, and different problems and their processing methods in self-adaptive auto-focusing of irregular images for the diffraction grating were discussed. A group of methods to control the auto-focusing of the irregular facula of the diffraction grating were proposed and a self-adaptive auto-focusing device were developed to achieve the offline detection of sampling distance and sampling angle for the diffraction grating. By optimizing threshold theory, the image was divided into two parts, facula area and background. On the basis of the mathematical morphology theory, the facula area, major and minor axis lengths and image center were calculated. Then, the adaptive focusing methods were implemented by analyzing the corresponding relationship between image center position and leveling motor, the lengths of major and minor axes and the axial motor. Finally, focusing iteration and centroid search were performed by combining the area minimization with the golden section algorithm. Test results show that the standard deviation of sampling distance is 412.5 μm and that of sampling angle is 10.35″, which meets the requirements for sampling distance and sampling angle less than 1000 μm and 30″, respectively. This device is characterized by simple structure, high reliability, strong anti-interference ability, and has been applied in a large aperture diffraction grating comprehensive diagnosis platform.Accession Number: 20142517852338