Robust moving object detection against fast illumination change

To solve the problem due to fast illumination change in a visual surveillance system, we propose a novel moving object detection algorithm for which we develop an illumination change model, a chromaticity difference model, and a brightness ratio model. When fast illumination change occurs, background pixels as well as moving object pixels are detected as foreground pixels. To separate detected foreground pixels into moving object pixels and false foreground pixels, we develop a chromaticity difference model and a brightness ratio model that estimates the intensity difference and intensity ratio of false foreground pixels, respectively. These models are based on the proposed illumination change model. Based on experimental results, the proposed method shows excellent performance under various illumination change conditions while operating in real-time.     

基于RGB色彩空间的四带树木图像颜色校正方法

针对图像受近红外光线参与的影响而出现偏色现象,介绍了一种基于RGB色彩空间的四带（RGB带和一个附加的近红外带）树木图像颜色校正方法. 该方法首先根据R-G差值图像,通过实验获取合适的阈值将四带树木图像二值化. 然后,利用中值滤波器和数学形态学对二值图像进行滤噪处理,提取出偏色较严重区域. 最后分区域进行颜色校正,达到消除近红外干扰的目的. 实验结果表明,该方法对部分常见含植物的场景图像有比较好的校正效果.     

基于LAAE网络的跨语言短文本情感分析方法

跨语言短文本情感分析作为自然语言处理领域的一项重要的任务, 近年来备受关注. 跨语言情感分析能够利用资源丰富的源语言标注数据对资源匮乏的目标语言数据进行情感分析, 建立语言之间的联系是该任务的核心.与传统的机器翻译建立联系方法相比, 迁移学习更胜一筹, 而高质量的跨语言文本向量则会提升迁移效果. 本文提出LAAE网络模型, 该模型通过长短记忆网络(LSTM)和对抗式自编码器(AAE)获得含上下文情感信息的跨语言向量, 然后利用双向GRU (Gated Recurrent Unite)进行后续情感分类任务. 其中, 分类器首先在源语言上进行训练,最后迁移到目标语言上进行分类任务. 本方法的有效性体现在实验结果中.     

Exploring a Stable and Dense 3D Lead Chloride Hybrid with Emission of Self-Trapped Excitons toward X-Ray Scintillation

The exceptional photophysical properties of 3D organic–inorganic lead halide hybrids (OILHs) endow their significant potential for usage in optoelectronics, which has sparked intense research on novel 3D OILHs and associated applications. However, constructing new 3D OILHs based on large organic cations suffers from tough challenges due to the limitation of the Goldschmidt tolerance factor rule, let alone further explorations of their practical applications. Herein, a brand-new 3D lead chloride hybrid, (1MPZ)Pb₄Cl₁₀·H₂O ( 1 , 1MPZ = 1-methylpiperazine) is reported, featuring a dense 3D lead chloride framework made of the corner-, edge-, and face-shared lead chloride polyhedra. 1 presents a broadband white light emission with a large Stokes shift and a nanosecond photoluminescence lifetime, which originates from radiative recombination of self-trapped excitons (STEs) induced by the highly distorted structure. Such a reabsorption-free and fast-decayed STEs emission coupling with the dense 3D architecture further enables 1 with effective X-ray scintillation with good sensitivity. Impressively, 1 also shows superior environmental and radiation stability. This study provides a new 3D OILH with appealing luminescence, not only expanding the 3D OILH family but also inspiring the exploitation of their optoelectronic applications.     

基于区域密度的交通拥堵判别方法

本文通过对交通流进行特征提取和状态的划分,建立了交通流判别准则.同时提出一种交通拥堵程度的智能判断方法,该方法以区域密度作为判别交通流状态的主要特征值,辅以时间占有率,结合空间和时间维度来判断当前道路的拥挤程度.仿真实验说明该方法具有较好的准确性和客观性.     

基于运动矢量交点密集度的人群恐慌行为检测

为了更准确有效的识别人群恐慌行为,本文提出了一种利用视频中人群运动矢量的交点密集度来判断人群恐慌异常的新算法.该算法以LK光流法为基础来提取运动人群的运动矢量信息,接着通过获得的信息求取运动矢量间的两两交叉点,然后运用分块法求得区域交叉点密集度,并以此来识别人群异常.对多个视频进行测试,测试结果表明:该算法能够以较高正确率识别视频中人群的恐慌行为.     

High capacity reversible data hiding scheme based upon discrete cosine transformation

In this paper, we propose a reversible data hiding scheme based on the varieties of coefficients of discrete cosine transformation of an image. Cover images are decomposed into several different frequencies, and the high-frequency parts are embedded with secret data. We use integer mapping to implement our 2-dimensional discrete cosine transformation. Thus, the image recovered from the modified coefficients can be transformed back to the correct data-hidden coefficients. Since the distribution of 2-dimensional DCT coefficients looks close to Gaussian distribution centralized at zero, it is a natural candidate for embedding secret data using the histogram shifting approach. Thus, our approach shifts the positive coefficients around zero to the right and the negative coefficients around zero to the left in order to leave a space to hide the secret data. The experimental comparisons show that, compared to Chang et al. and Lin et al.'s method, the embedding capacity and quality of the stego-image of the proposed method is a great improvement.     

Carrier dynamics and intervalley scattering in InN

The carrier cooling and the carrier relaxation of an InN thin film illuminated with two excitation energies of 1.53 and 3.06 eV were studied by an ultrafast time-resolved photoluminescence upconversion apparatus. The hot phonon effect could be accounted for longer effective phonon emission times as compared to the theoretical prediction. The rise time and the LO phonon emission time for 3.06 eV excitation were much smaller than those for 1.53 eV excitation. These differences were attributed to the intervalley scattering between the Γ₁ and Γ₃ valleys in InN when carriers were excited with the energy of 3.06 eV. The intervalley scattering times of 250 fs and 2 ps were estimated for the intervalley scattering from the Γ₁ to Γ₃ valley and the reversed scattering process, respectively.     

Light absorption distribution of prostate tissue irradiated by diffusing light source

Determination of light absorption distribution in the prostate tissue irradiated by diffusing light source is important for the treatment planning.In this paper,a three-dimensional(3D)optical model of human prostate is developed,and the light absorption distribution in the prostate tissue is estimated by Monte Carlo simulation method.Light distribution patterns including 3D distributions in the tissue model irradiated by two diffusing light sources are obtained and compared.Also,the impacts of length and energy of cylinder diffusing light source on the irradiance volume are demonstrated.Those results will be significant for the nondestructive qualitative assessments of photodosimetry in biomedical phototherapy.     

利用串联双滤波技术改善声光滤波器光谱分辨率的实验研究

考虑旋光性设计 制作了两个非共线型声光可调滤波器(AOTF),通过前后放置对信号光进行两次的级联滤 波。 调节前后两个AOTF所加的超声频率值, 保证前后两个AOTF存在透射光谱的叠加区,从而实现光谱分辨率的提升及调控。实 验结果表明,双滤波技术还可以有效抑制AOTF透射光谱的旁瓣效应,改善透射光 谱的纯度。