深度学习算法下多模态人体动作实时跟踪仿真

人体的运动姿态多样,光影和噪声等因素对人体运动特征的提取和分析具有很大影响。且人体动作跟踪过程若目标存在遮挡,易导致目标丢失问题,严重影响人体动作跟踪的精准度。为解决上述问题,引入深度学习方法中的自编码器,提出新的多模态人体动作实时跟踪方法。通过对人体动作图像采集处理,完成人体动作特征的粗提取;利用自编码器对粗特征增强,实现人体动作特征的细提取,建立常规人体动作特征模板;利用核相关滤波(kernel correlation filter,KCF)获取响应函数峰值,建立目标丢失判据,结合特征匹配算法,计算跟踪目标与区域化结构特征之间相似度,确定最佳跟踪目标,实现人体动作的实时跟踪。实验结果表明,所提方法的动作跟踪精准度高,时延短,表明所提方法的跟踪效果好。

Simulation of Atmospheric Wave-Fronts with Turbulence Intermittency

A new extendable method for the simulation of atmospheric wave-fronts with turbulence intermittency is reported. The purpose is to generate simulations consistent with the distributions observed for the turbulence parameters and the seeing, not available with standard methods. The intermittency is included by entering log-normal distributed arrays for the Fried parameter and the spatial coherence outer scale length into an extended form of the phase spectrum. The method is tested on large samples of simulated long-exposure point-source images. The tests show the agreement of the simulations with literature data. The simulations show that the intermittency affects negligibly the long-term median image size but breaks the symmetry of the wave-front phase spectrum, scatters the phase structure function and changes the image profile.

基于连续帧图像面积变化的火灾探测方法

针对传统视频火灾探测方法误报率高、适用性弱的缺点,基于火焰在连续帧图像上的面积变化规律,提出了一种新的火灾探测方法。首先利用RGB-HSI颜色模型筛选出单帧图像中的疑似火焰区域,再结合区域跟踪算法和面积极大值统计计算出疑似火焰区域的闪烁频率,从而实现火灾探测。酒精火和远距离强光背景下的火焰探测实验证明:该方法能够排除常见干扰源的影响,对于室内火灾探测具有一定的实用价值。

我国南极科考站附近气旋的特征分析

本文基于欧洲中心的ERA-Interim再分析数据,采用雷丁大学的气旋自动识别与追踪算法,建立了1979-2013年中国南极科学考察站长城站、中山站以及泰山站划定的统计区域内气旋数据集。基于这套数据,分析了3个站区附近气旋的气候特征,包括:气旋数量、空间分布、强度、加深率等,并研究了达到爆发性发展的气旋的时空特征。分析表明,长城站以及中山站气旋数量均有增加趋势,但都未达到显著水平;泰山站气旋数量的减少趋势达到显著性水平。3个站区的气旋统计表明,长城站统计区域内气旋活动最频繁,且爆发性气旋活动相对较多;中山站、长城站气旋的数量和空间分布具有季节变化,气旋数量夏季较其他3个季节偏少、偏弱;泰山站气旋活动最少,并且大部分影响到该站的气旋都处于消亡阶段,气旋的强度较弱,故气旋活动对该站的影响较弱。

GLOBAL EXISTENCE OF SOLUTIONS FOR A MULTI-PHASE FLOW: A BUBBLE IN A LIQUID TUBE AND RELATED CASES

In this paper we study the problem of the global existence （in time） of weak entropic solutions to a system of three hyperbolic conservation laws, in one space dimension, for large initial data. The system models the dynamics of phase transitions in an isothermal fluid; in Lagrangian coordinates, the phase interfaces are represented as stationary contact discontinuities. We focus on the persistence of solutions consisting in three bulk phases separated by two interfaces. Under some stability conditions on the phase configuration and by a suitable front tracking algorithm we show that, if the BV-norm of the initial data is less than an explicit （large） threshold, then the Cauchy problem has global solutions.

Variations in Extratropical Cyclone Activity in Northern East Asia

Based on an improved objective cyclone detection and tracking algorithm, decadal variations in extratropical cyclones in northern East Asia are studied by using the ECMWF 40 Year Reanalysis （ERA-40） sea-level pressure data during 1958-2001. The results reveal that extratropieal cyclone activity has displayed clear seasonal, interannual, and decadal variability in northern East Asia. Spring is the season when cyclones occur most frequently. The spatial distribution of extratropieal cyclones shows that cyclones occur mainly within the 40°- 50°N latitudinal band in northern East Asia, and the most frequent region of occurrence is in Mongolia. Furthermore, this study also reveals the fact that the frequency of extratropieal cyclones has significantly decreased in the lower latitude region of northern East Asia during 1958 2001, but deeadal variability has dominated in higher latitude bands, with frequent cyclone genesis. The intensity of extratropical cyclones has decreased on an annual and seasonal basis. Variation of the annual number of cyclones in northern East Asia is associated with the mean intensity of the baroelinie frontal zone, which is influenced by climate warming in the higher latitudes. Moreover, the dipole structure of extratopical cyclone change, with increases in the north and decreases in the southern part of northern East Asia, is related to the northward movement of the baroelinic frontal zone on either side of 110°E.

Measurements of AIM for Visible Wavelength Based on Individual Eye Model

We measure and calculate the aerial image modulation （AIM） of human retina for visible wavelengths based on the individual eye model. By employing the optical design software ZEMAX, we obtain the modulation transfer function （MTF） of human eye in visible wavelengths. Using CSV-1000 and VAF-1000, the contrast sensitivity function （CSF） and the visual acuity （VA） for the same eye are measured. Then the AIM of human retina could be acquired by the relations between MTF and CSF and between MTF and VA. The AIM of human retina is independent of MTF, and the values of AIM for normal eyes （without retina disease） are similar, so the assembly average for large numbers of normal eyes can be a standard AIM curve, which is helpful for the diagnosis of diseases in the retina system.

Generation of orbital angular momentum and focused beams with tri-layer medium metamaterial

We propose a metal/dielectric tri-layer metamaterial for wavefront shaping. By arranging the element in an array with a constant phase gradient and irradiated it with a plane wave, focused and focused vortex beams can be obtained. The designed metamaterial features the excellent capability of focused/focused vortex beams generation within the operating frequency range of 30 GHz–34 GHz. The simulation results are consistent with the theoretical analyses.

Extending Dynamic Range of Phase Diversity

A method for extending dynamic range of phase diversity （PD） is proposed.The estimating of pupil phase is divided into two steps.Low-order Zernike coefficients of the phase are estimated firstly before all of the Zernike coefficients estimated.The method is based on the assumption that the aberrations of the system are dominated by low order Zernike modes,which is true to most practical situations.The algorithm of BroydenFletcher-Goldfarb-Shanno（BFGS） is used in the two steps for optimizing process.Computer simulations are conducted after the theoretical analysis.The result shows that this method can effectively avoid the trap in local minimum during iteration process,and the dynamic range that PD can detect extends from about 2λ to about 4λ.

Wave-front correction of high-intensity fs laser beams by using closed-loop adaptive optics system

We developed an adaptive optics system to correct the wave-front distortion of an intense fs laser beam from our multi-TW laser system, Jiguang II. In this paper, the instruments of the adaptive optical system are described and the experimental results of the closed-loop wave-front correction are presented. A distorted laser wave-front of 20 wavelengths of P-V values was corrected to 0.15 wavelength of P-V values. The beam quality of the laser system varies from 3.5 diffraction limit to 1.5 diffraction limit.

Focal plane wave-front sensing algorithm for high-contrast imaging

High-contrast imaging provided by a coronagraph is critical for the direction imaging of the Earth-like planet orbiting its bright parent star.A major limitation for such direct imaging is the speckle noise that is induced from the wave-front error of an optical system.We derive an algorithm for the wave-front measurement directly from 3 focal plane images.The 3 images are achieved through a deformable mirror to provide specific phases for the optics system.We introduce an extra amplitude modulation on one deformable mirror configuration to create an uncorrelated wave-front,which is a critical procedure for wave-front sensing.The simulation shows that the reconstructed wave-front is consistent with the original wave-front theoretically,which indicates that such an algorithm is a promising technique for the wave-front measurement for the high-contrast imaging.

Optical study on the vision correction and supernormal vision based on the wave-front aberrations of human eye

In this paper we present the recent research results in the field of vision correction and supernormal vision according to the actual measurements of the wave-front aberrations and the corneal surface topography,the clinical detection of the visual function and the laser corneal refractive surgery,and the optimization of the optical system. These include the features of the aberrations of human eye with different pupil sizes,different fields of view and temporal accommodation,the influence of the polychromatic illumination of the visible wavelength on the supernormal vision,and the effect of the existing laser corneal refractive surgery on the wave-front ab-errations of the eye. It is shown that the wave-front aberration of human eye is of temporal variation and of synthesis with multi impact factors. To achieve super-normal vision,an optimum engineering data for the customized laser corneal sur-gery should be firstly acquired,which may involve the dynamic free-form optical surface. Although the myopia can be corrected by the laser in situ keratomileusis(LASIK) in a certain degree,it brings about negative effects under scotopic condi-tions.

Principle demonstration of fine pointing control system for inter-satellite laser communication

Due to high data rates and reliability,inter-satellite laser communication has developed rapidly in these days.However,the stability of the laser beam pointing is still a key technique which needs to be solved;otherwise,the beam pointing jitter noise would reduce the communication quality or,even worse,would make the inter-satellite laser communication impossible.For this purpose,a bench-top of the fine beam pointing control system has been built and tested for inter-satellite laser communication.The pointing offset of more than 100rad is produced by the steering mirror.With beam pointing control system turned on,the offset could be rapidly suppressed to lower than 100 nrad in less than 0.5 s.Moreover,the pointing stability can be kept at 40 nrad for yaw motion and 62 nrad for pitch motion,when the received beam jitter is set at 20rad.

A non-stationary channel model for 5G massive MIMO systems

We propose a novel channel model for massive multiple-input multiple-out （MIMO） communication systems that incorporate the spherical wave-front assumption and non-stationary properties of clusters on both the array and time axes. Because of the large dimension of the antenna array in massive MIMO systems, the spherical wave-front is assumed to characterize near-field effects resulting in angle of arrival （AoA） shifts and Doppler frequency variations on the antenna array. Additionally, a novel visibility region method is proposed to capture the non-stationary properties of clusters at the receiver side. Combined with the birth-death process, a novel cluster evolution algorithm is proposed. The impacts of cluster evolution and the spherical wave-front assumption on the statistical properties of the channel model are investigated. Meanwhile, corresponding to the theoretical model, a simulation model with a finite number of rays that capture channel characteristics as accurately as possible is proposed. Finally, numerical analysis shows that our proposed non-stationary channel model is effective in capturing the characteristics of a massive MIMO channel.

Revisiting the Parallel Strategy for DOACROSS Loops

DOACROSS loops are significant parts in many important scientific and engineering applications,which are generally exploited pipeline/wave-front parallelism by loop transformations.However,previous work almost statically performs iterations in parallel threads,thus causing a waste of computing resources in thread synchronization.This paper proposes a brand-new parallel strategy for DOACROSS loops that provides a dynamic task assignment with reduced dependences to achieve wave-front parallelism through loop tiling.The proposed strategy uses a master-slave parallel mode and some customized structures to realize dynamic and flexible parallelization,which effectively avoids threads from waiting in communication.An efficient tile size selection(TSS)approach is also proposed to preserve data reuse in cache for tiled codes.The experimental results show that the proposed parallel strategy obtains good and stable speedups over six typical benchmarks with different problem sizes and different numbers of threads on an Intel■Xeon■32-core serve?.And it outperforms two static strategies,a barrier-based strategy and a post/wait-based strategy,by 32% and 20% in average performance,respectively.This strategy also yields a better performance than a mutex-based dynamic strategy.Besides,it has been demonstrated that the proposed TSS approach can achieve a near-optimal performance and is comparable with a state-of-the-art TSS approach.