Sentinel-2卫星的多光谱轻量级船舶目标检测算法

近年来深度卷积神经网络在可见光船舶检测方面取得了显著的进展，然而，大多数相关研究是通过改进大型的网络结构来提高检测性能，因此加大了对更高计算机性能的需求。此外，可见光图像难以在云、雾、海杂波、黑夜等复杂场景检测到船舶。针对以上问题，提出了一种融合红（red, R）、绿（green, G）、蓝（blue, B）和近红外（NIR）4个波段光谱信息的由粗到精细的轻量型船舶检测算法。与现有的方法中根据光谱特性利用水体检测算法提取水体区域不同之处是该算法是利用改进的水体检测算法来提取船舶候选区域。为获取更准确的候选区域，对船舶、厚云、薄云、平静海面、杂波海面5种场景中4个波段的像素值进行了统计分析，选取近红外大于阈值作为辅助判断，并以其中心点获取候选区域32×32大小的切片，并对切片进行非极大值抑制，由此获得了船舶粗检测结果。随后构建了轻量级LSGFNet网络对船舶候选区域切片进行精细识别。构建的网络融合了1×1卷积提取的波谱特征与3×3的提取几何特征，为防止光谱特征与几何特征的信息在融合时“信息不流通”，在LSGFNet网络中引入了ShuffleNet中的通道打乱机制，并减小了模型结构，与典型的轻量级网络相比具有更好的效果且模型较小。最后，利用Sentinel-2卫星多光谱10 m分辨率数据构建了512×512大小的1 120组数据进行粗检测，以及32×32大小的6 014组数据进行精细网络训练，其中候选区域粗提取的查全率为98.99%，精细识别网络精确度为96.04%，不同场景下的平均精确度为92.98%。实验表明该算法在抑制云层、海浪杂波等干扰的复杂背景下具有较高的检测效率，且训练时间短、计算机性能需求低。     

Collective Tunnelling of Strongly Interacting Cold Atoms in a Double-Well Potential

It is known that under resonance conditions, a group of strongly interacting bosonic atoms, trapped in a double-well potential, mimics a single particle, performing Rabi oscillations between the wells. By implication, all atoms need to tunnel at roughly the same time, even though the Bose–Hubbard Hamiltonian accounts only for one-atom-at-a-time transfers. The mechanism of this collective behavior is analyzed, the Rabi frequencies in the process are evaluated, and the limitation of this simple picture is discussed. In particular, it is shown that the small rapid oscillations superimposed on the slow Rabi cycle result from splitting the transferred cluster at the sudden onset of tunnelling, and disappear if tunnelling is turned on gradually.     

TBOR-shaped electro-optic modulator with dual output based on double-layer graphene

We proposed an electro-optic modulator with two-bus one-ring (TBOR) structure to improve the extinction ratio and reduce insert loss. It has a dual output compared with one-bus one-ring structure. In addition, double-layer graphene makes it possible for the modulation in the visible to mid-infrared wavelength range. It shows that this new electro-optic modulator can present two switching states well with low insertion loss, high absorption and high extinction ratio. At $\lambda =1550\text{ nm}$, when the switching states are based on the chemical potential, ${\mu }_{\mathrm{c}}=0.38\text{ eV}$ and ${\mu }_{\mathrm{c}}=0.4\text{ eV}$, the insertion losses of both output ports are less than 2 dB, the absorption of the output port coupled via a micro-ring reaches 45 dB and the extinction ratio reaches 14 dB. When the refractive index of the dielectric material is 4.2, the applied voltage will be less than 1.2 V, thus can be used in low-voltage CMOS technology.     

Comparative study on transport properties of N-, P-, and As-doped SiC nanowires: Calculated based on first principles

According to the one-dimensional quantum state distribution, carrier scattering, and fixed range hopping model, the structural stability and electron transport properties of N-, P-, and As-doped SiC nanowires(N-SiCNWs, P-SiCNWs, and As-SiCNWs) are simulated by using the first principles calculations. The results show that the lattice structure of NSiCNWs is the most stable in the lattice structures of the above three kinds of doped SiCNWs. At room temperature,for unpassivated SiCNWs, the doping effect of P and As are better than that of N. After passivation, the conductivities of all doped SiCNWs increase by approximately two orders of magnitude. The N-SiCNW has the lowest conductivity. In addition, the N-, P-, As-doped SiCNWs before and after passivation have the same conductivity–temperature characteristics,that is, above room temperature, the conductivity values of the doped SiCNWs all increase with temperature increasing.These results contribute to the electronic application of nanodevices.     

Experimental and Theoretical Mechanistic Study on the Thermal Decomposition of 3,3-diphenyl-4-(trichloromethyl)-5-nitropyrazoline

The present paper is a continuation of comprehensive study regarding to synthesis and properties of pyrazoles and their derivatives. In its framework an experimental and theoretical studies of thermal decomposition of the 3,3-diphenyl-4-(trichloromethyl)-5-nitropyrazoline were performed. It was found, that the decompositions of the mentioned pyrazoline system in the solution and at the melted state proceed via completely different molecular mechanisms. These mechanisms have been explained in the framework of the Molecular Electron Density Theory (MEDT) with the computational level of B3LYP/6-31G(d). A Bonding Evolution Theory (BET) examination of dehydrochlorination of the 3,3-diphenyl-4-(trichloromethyl)-5-nitropyrazoline permits elucidation of the molecular mechanism. It was found, that on the contrary for most known HCl extrusion processes in solution, this reaction is realised via single-step mechanism.     

A new study of associating inhomogeneous fluids with classical density functional theory

ABSTRACT

A new density functional for the study of associating inhomogeneous fluids based on Wertheim's first-order thermodynamic perturbation theory is presented and compared to the most currently used associating density functionals. This functional is developed using the weighted density approximation in the range of association of hard spheres. We implement this functional within the framework of classical density functional theory together with modified fundamental measure theory to account for volume exclusion of hard spheres. This approach is tested against molecular simulations from literature of pure associating hard spheres and mixtures of non-associationg and associating hard spheres with different number of bonding sites close to a hard uniform wall. Furthermore, we compare and review our results with the performance of associating functionals from literature, one based on fundamental measure theory and the inhomogeneous version of Wertheim's perturbation theory. Results obtained with classical DFT and the three functionals show excellent agreement with molecular simulations in systems with one hard wall. For the cases of small pores where only one or two layers of fluid are allowed discrepancies between results with classical DFT and molecular simulations were found.     

Spectral estimates for saddle point matrices arising in weak constraint four‐dimensional variational data assimilation

We consider the large sparse symmetric linear systems of equations that arise in the solution of weak constraint four‐dimensional variational data assimilation, a method of high interest for numerical weather prediction. These systems can be written as saddle point systems with a 3 × 3 block structure but block eliminations can be performed to reduce them to saddle point systems with a 2 × 2 block structure, or further to symmetric positive definite systems. In this article, we analyse how sensitive the spectra of these matrices are to the number of observations of the underlying dynamical system. We also obtain bounds on the eigenvalues of the matrices. Numerical experiments are used to confirm the theoretical analysis and bounds.     

基于时变Copula模型的 基金收益率相依关系的应用研究

基于时变Copula模型,获得预测方差,确定单个基金收益率序列的边缘分布.利用常见的静态Copula和时变Copula模型对基金收益率序列间两两相依关系进行建模并进行对比分析.应用研究表明,基于MCMC方法的时变Copula模型能更有效地度量基金收益率序列的风险.