Analysis and modeling of radiometric error caused by imaging blur in optical remote sensing systems

Imaging blur changes the digital output values of imaging systems. It leads to radiometric errors when the system is used for measurement. In this paper, we focus on the radiometric error due to imaging blur in remote sensing imaging systems. First, in accordance with the radiometric response calibration of imaging systems, we provide a theoretical analysis on the evaluation standard of radiometric errors caused by imaging blur. Then, we build a radiometric error model for imaging blur based on the natural stochastic fractal characteristics of remote sensing images. Finally, we verify the model by simulations and physical defocus experiments. The simulation results show that the modeling estimation result approaches to the simulation computation. The maximum difference of relative MSE (Mean Squared Error) between simulation computation and modeling estimation can achieve 1.6%. The physical experimental results show that the maximum difference of relative MSE between experimental results and modeling estimation is only 1.29% under experimental conditions. Simulations and experiments demonstrate that the proposed model is correct, which can be used to estimate the radiometric error caused by imaging blur in remote sensing images. This research is of great importance for radiometric measurement system evaluation and application.     

水声时变线谱建模与双稳态随机共振分析

建立了水声时变线谱的模型,将双稳态随机共振系统应用于水声时变线谱信号的检测,考查了随机共振系统对水声环境的适应能力,提出了设计水声线谱检测系统可以利用的外在参数。在信噪比变化、声压起伏、线谱漂移等情况下,仿真结果表明双稳态随机共振系统都能较好地工作。为水声领域新型线谱检测系统的设计提供了依据。     

Stochastic bifurcation in FitzHugh-Nagumo ensembles subjected to additive and/or multiplicative noises

We have studied the dynamical properties of finite N-unit FitzHugh-Nagumo (FN) ensembles subjected to additive and/or multiplicative noises, reformulating the augmented moment method (AMM) with the Fokker-Planck equation (FPE) method [H. Hasegawa, J. Phys. Soc. Japan 75 (2006) 033001]. In the AMM, original 2N-dimensional stochastic equations are transformed to eight-dimensional deterministic ones, and the dynamics is described in terms of averages and fluctuations of local and global variables. The stochastic bifurcation is discussed by a linear stability analysis of the deterministic AMM equations. The bifurcation transition diagram of multiplicative noise is rather different from that of additive noise: the former has the wider oscillating region than the latter. The synchronization in globally-coupled FN ensembles is also investigated. Results of the AMM are in good agreement with those of direct simulations (DSs).     

Power-law distribution of gene expression fluctuations

Large-scale genomic technologies has opened new possibilities to infer gene regulatory networks from time series data. Here, we investigate the relationship between the dynamic information of gene expression in time series and the underlying network structure. First, our results show that the distribution of gene expression fluctuations (i.e., standard deviation) follows a power-law. This finding indicates that while most genes exhibit a relatively low variation in expression level, a few genes are revealed as highly variable genes. Second, we propose a stochastic model that explains the emergence of this power-law behavior. The model derives a relationship that connects the standard deviation (variance) of each node to its degree. In particular, it allows us to identify a global property of the underlying genetic regulatory network, such as the degree exponent, by only computing dynamic information. This result not only offers an interesting link to explore the topology of real systems without knowing the real structure but also supports earlier findings showing that gene networks may follow a scale-free distribution.     

Stability analysis of impulsive stochastic Cohen-Grossberg neural networks with mixed time delays

In this paper, the problem of stability analysis for a class of impulsive stochastic Cohen-Grossberg neural networks with mixed delays is considered. The mixed time delays comprise both the time-varying and infinite distributed delays. By employing a combination of the M-matrix theory and stochastic analysis technique, a sufficient condition is obtained to ensure the existence, uniqueness, and exponential p-stability of the equilibrium point for the addressed impulsive stochastic Cohen-Grossberg neural network with mixed delays. The proposed method, which does not make use of the Lyapunov functional, is shown to be simple yet effective for analyzing the stability of impulsive or stochastic neural networks with variable and/or distributed delays. We then extend our main results to the case where the parameters contain interval uncertainties. Moreover, the exponential convergence rate index is estimated, which depends on the system parameters. An example is given to show the effectiveness of the obtained results.     

有机环境污染物紫外光谱检索的神经网络方法

详细讨论了网络优化参数、模拟的测量过程中噪声及杂质对网络收敛性能及预测误差的影响。为加速网络收敛,提高紫外光谱检索的正确率,采用了导数光谱对反向传播的人工神经网络(BP-ANN)进行训练和检索,该方法对检索光谱中噪声、杂质,尤其是斜坡背景的允许程度明显提高。文章还将ANN方法与普通的相关系数法的识别结果进行了比较。结果表明,优化参数下的人工神经网络的库检索法在抗噪、容杂等方面都明显地优于普通的相关系数法,是一种很有效的紫外库检索方法。     

基于CouchDB和Elastic Search的高性能化学结构搜索引擎与数据库的构建

计算机辅助的化学结构搜索在化学信息学中地位十分重要,本文设计了一套高性能的化学结构和化学数据搜索系统,称为DCAIKU.DCAIKU基于CouchDB无模式数据库和ElasticSearch基础架构构建,通过将结构相似性搜索变换为文字搜索实现了高性能和高灵活性的检索引擎：在满足化学信息存储的高灵活性条件下,仍然可以做到低延迟和高准确性,同时拥有良好的伸缩性,可以大规模并行化和集群化.     

自适应光学系统几种随机并行优化控制算法比较

 直接对系统性能指标进行优化是自适应光学系统中一种重要的波前畸变校正方法,选择合适的随机并行优化控制算法是该技术成功实现的关键。以32单元变形镜为校正器,基于多种随机并行优化算法建立自适应光学系统仿真模型。从算法的收敛速度、校正效果、局部极值3个方面对遗传算法、单向扰动随机并行梯度下降、双向扰动随机并行梯度下降及模拟退火算法进行了比较。仿真结果表明,遗传算法收敛速度太慢,不适用于需要实时控制的自适应光学系统;双向扰动随机并行梯度下降算法收敛速度、校正效果要优于单向扰动随机并行梯度下降,且能够适应各种情况下的扰动电压;模拟退火几乎以概率1收敛到全局极值附近,且收敛速度是上述算法中最快的。     

Stochastic resonance characteristic analysis of new potential function under Levy noise and bearing fault detection

Based on the output saturation of classcial bistable stochastic resonance (CBSR), a new type of piecewise nonlinear bistable stochastic resonance (PNBSR) system is constructed. The mean signal-to-noise ratio gain is regarded as an index to measure the stochastic resonance phenomenon. The laws for the resonant output of piecewise nonlinear bistable system governed by l, c, a, b and D of Levy noise are explored under different characteristic index α and symmetry parameter β of Levy noise. The results show that the output of PNBSR system has increased 4?dB by comparing with the output signal-to-noise ratio of CBSR system. And the stochastic resonance phenomenon can be induced by adjusting the piecewise nonlinear system's parameters under any α or β of Levy noise. The interval of the parameters of system which induces good stochastic resonance is roughly the same. And the output signal waveform of resonance is very similar to the input signal waveform, which has some reference value for the signal recovery. Moreover, we can find the good stochastic resonance interval of the system parameters do not change with D of Levy noise under the different noise intensity D of Levy noise. On the basis of this, adjusting the intensity amplification factor D of Levy noise, which induces good stochastic resonance, and the interval does not change with α or β. At last, the piecewise nonlinear bistable system is applied to detect bearing fault signals, which achieves better performance compared with the classical bistable system.     

Estimating the mixing time of concert halls using the eXtensible Fourier Transform

In room acoustics, we measure room impulse responses (RIRs) in order to fully describe the hall. RIRs are composed of a succession of arrivals (i.e., some sound rays which have undergone one or more reflections on their way from the source to the receiver). We propose the eXtensible Fourier Transform (XFT) in order to investigate the time evolution of spectral components of RIRs. The phase evolution versus time allows to estimate the mixing time, which is defined as the time it takes for initially adjacent sound rays to spread uniformly across the room. After presenting some properties of the XFT, we show why one must compensate the natural energy decay of the RIR in order to obtain stationary signals. We estimate mixing times for a set of experimental RIRs and several that are synthesized using a stochastic model. Then, we estimate the dependance of mixing time upon the source/receiver distance in all these RIRs. Results are consistent up to the lack of reproducibility of the sound sources, but are strongly dependent on some parameters used for computing the XFT. We finally discuss the relevance of the name mixing time with respect to the theory and in regard to the time we estimate, that we propose to call cross-over time.