Abstract:A new stopband filter consisted of dielectric grating waveguide was proposed and its filter characteristics were carefully examined by an approach of combining the muhimode network theory with the rigorous mode-matching method. The Brillouin diagram for the dominant TM mode and the variations of the normalized center frequency and the band width as well as the maximum attenuation constant of the stopband with the geometry parameters were presented. By comparing with the traditional dielectric grating stopband filter, it has been found that the new filter structure not only has the advantage of lower conductor loss, but also is of larger stopband bandwidth and maximum attenuation in the band than those of the traditional ones. The numerical results presented here are of guiding significance for the design of new band-rejected filters.

Abstract:通过对不同组分的铌镁钛酸铅(PMNT)薄膜的紫外-可见透射光谱以及红外椭圆偏振光谱的分析,利用Tauc-Lorentz(T-L) Cauchy色散关系和经典中红外色散关系,得到了该材料在可见和中红外区的光学常数,发现在可见和中红外区,薄膜的折射率随着PT(PbTiO3)含量的增加而增大,但是薄膜的光学禁带宽度随之而减小.

Abstract:实现红外焦平面阵列非均匀性自适应校正是高级红外探测系统追求的重要目标,对提高红外探测系统的空间分辨率、温度分辨率、探测距离以及辐射量的正确度量具有重要意义.归纳总结了国内外关于凝视红外焦平面阵列非均匀性自适应校正算法的部分研究工作及其进展,比较了典型自适应算法的性能和适用条件,为进一步开展红外焦平面阵列非均匀性自适应校正研究提供参考意见.

Abstract:为了切断疯牛病的传播途径保证饲料安全,开展非反刍动物肉骨粉中反刍动物成分含量的检测方法研究十分必要.本文探讨了利用近红外漫反射光谱(MRS)分析技术快速检测非反刍动物(猪鸡)肉骨粉中反刍动物(牛羊)成分含量的可行性.收集了不同地区不同种类的肉骨粉样本,在猪和鸡肉骨粉中分别掺入0～48 %的牛或羊肉骨粉,制备了200个样本.针对含牛羊源的猪肉骨粉、含牛羊源的鸡肉骨粉和含牛羊源的猪鸡肉骨粉三类样本,以掺入的牛羊源成分的质量比例作为真值,分别采用改进的偏最小二乘回归(MPLS)、偏最小二乘回归(PLS)和主成分回归(PCR)三种不同建模方法建立了MRS定标模型.用独立的验证集对模型进行了验证,最优模型验证集的决定系数(R2)均大于0.90,相对分析误差(RPD)均大于3.0.结果表明,利用NIRS分析技术可以快速检测出猪鸡肉骨粉中牛羊源成分的含量.

Abstract:Aiming at the characteristics of low visible light images and infrared images, a novel adaptive image fusion scheme based on DT-CWT was presented. The technique has good shift-invariance and directional selectivity, and is more suitable for human vision. The DT-CWT was firstly used to perform a muhiresolution decomposition of source images. By taking advantage of the characters of the coefficients, the immune clonal selection algorithm was introduced in low-pass subbands and almost optimal fused weights were obtained adaptively. To high-pass subbands, the local directive contrast was defined, which was based on human visual system. And then the contrast of fused images was enhanced and the detail information of source images was protected. The experimental results show that our fused technique is effective and the fused images have a better visual quality than their wavelet counterparts.

Abstract:波段选择是导弹助推段天基红外探测系统设计首先需要解决的问题.无论是中心波长的确定,还是波段宽度的设定,都应有合理的依据.本文提出了导弹助推段天基探测的像面照度模型,并根据模型得到导弹助推段天基红外探测系统的目标与背景信杂比作为目标函数,利用信杂比的最大化作为波段宽度选择的依据,进行探测波段的选择.分析结果表明,以H2O,CO2吸收带2.7μm和4.3μm为作为红外探测系统的中心波长有其固有的合理性,不同的波段宽度选择将使信杂比产生显著变化.本文所建立的像面照度模型对遥感器设计和性能估计亦有重要意义.

Abstract:For target rotation in IR image sequence, an improved Mean shift tracking method was approached. Mean shift vector was established by Gaussian kernel with bandwidth matrix, and recursive equations were given in EM like manner. Flexibility of algorithm was improved by motion compensation of image sequence. Robustness of algorithm was improved by combining with Kalman filter. Experiments show that our method achieves success in IR target tracking, at same time the method can be adapted to the transformations of rotation and zooming.

Abstract:建立了一种基于独立组分分析的可见/近红外光谱反射技术快速鉴别稻谷年份的新方法.首先用独立组分分析方法获取不同年份稻谷的可见/近红外光谱载荷图,将载荷图中相关性最大的波段(特征波段)作为人工神经网络的输入建立稻谷年份的鉴别模型.每个年份40个样本,3个年份共120个样本用来建立BP神经网络模型,剩余的3个年份各20个样本用于预测.预测的结果表明,准确率达到100%.同时通过独立组分分析,得到了稻谷主要成分对应的敏感波段.说明本文提出的基于独立组分分析的方法具有很好的鉴别效果,为稻谷的年份鉴别提供了一种新方法.

Abstract:对红外序列图像中的人体检测问题进行了研究,提出一种新的人体检测方法.首先采用自适应高斯混合模型对序列图像中背景进行建模,在准确分割出前景运动目标的基础上,提出了一种新的人体形状表达模型,充分考虑了多个人体发生粘连或互相遮挡的情况,并用亮度投影的方法对其进行分离;以人体表达模型作为输入向量,构建支持向量机(SVM,Support Vector Machine)对人体进行分类判别.不同红外视频序列的检测结果表明了所提出算法在单个人体和多人体情况下均具有较好的鲁棒性和可行性.

Abstract:研究了一种将变换矩阵分解为三角元素可逆矩阵(TERM)实现的主成分变换整数近似算法(IPCT).为限制误差和提高计算效率,改进了TERM分解中选择主元的方法.结合可逆整数PCT和三维Tarp编码技术,提出了一种新的高光谱图像无损压缩算法.该算法在进行空间维小波变换以后,利用改进的IPCT去除谱间相关;在编码阶段,新颖的三维Tarp编码器能利用五个简单的递归滤波器进行概率估计,以驱动一个非自适应算术编码器,对变换系数的显著性映射和细化信息进行熵编码.该算法复杂度较低,能够产生嵌入式码流,并且较已有的算法能获得更高的压缩比.

Abstract:By considering the urban entironment effect and remote sensing interpretation ability synthetically, the urban entironment basal status was classified into eight classes. Taking the Shanghai within outer line as a case, information extraction methods of urban entironment basal status from remote sensing images were studied. By taking TM and SPOT fusion image as the remote sensing data sources, and based on the knowledge of eight classes identification from high resolution remote sensing images with eyes in the GIS database, classification and regression tree recognition methods were built for different types of urban entironment basal status from the fused images and the fusion data ratio images. Four classes information including residential areas, water bodies, urban virescence and agriculture were extracted from the remote sensing data for a large area of city.

Abstract:方向信息提取与小样本问题是遥感目标识别与应用的瓶颈,基于平稳小波的Contourlet变换(Stationary Wavelet Based Contourlet Transform,简称SWBCT)与投影特征相结合,本文提出了一种新的遥感目标特征提取与识别方法.首先采用改进的Contourlet变换SWBCT对遥感图像进行分解,然后对分解子带进行方向投影并求投影向量的统计特性作为目标特征,进而利用K近邻分类器进行目标识别.与小波、Contourlet等代表性方法进行比较,实验结果与分析表明,该方法不仅具有较高的识别率,而且在小样本情况下,依然可以得到较好的结果.

Abstract:Detecting desertification by using remote sensing is an important task for China and other semi - arid and arid areas in the world. In this study, MODIS 1 km NDVI time series data of 2002 were used to classify land cover type in Xinjiang province and surrounding regions. During the classification, sparse areas, which are the areas at risk of desertification, are emphasized. A classification method, which is adapted to various land cover type and makes sparse vegetation emphasis, is introduced. This method better synthesizes seasonal influence factors and variable factors of natural conditions. Research results indicate that 0.24 million km^2 in Xinjiang and surrounding regions are potential desertification regions.

Abstract:Unlike the traditional direct current millimeter-wave radiometric imaging, an alternating current radiometer was first presented for the passive millimeter-wave imaging with simple and integrated structure in China. On basis of discussing the special structure and energy spectrum, the energy compensation mechanism and image description of an alternating current radiometric imaging were also analyzed. Lots of indoor and outdoor experiments were carried out for the different application such as safety inspection. The imaging results of 3mm band alternating current radiometer show that the alternating current radiometric image possesses its own characteristic different from the direct current radiometer. The alternating current radiometer is easy to realize passive millimeter-wave multi-channels focal plane imaging.

Abstract:A new method to calibrate CCD camera was proposed. It consists of two-dimension rotate instrument, collimator and CCD camera. At first, through Hough translation and least squares image matching, the intersection point of crisscross fixed on collimator, which imaged on the CCD focus plane, was found, And then a linear regression of least squares was used to calibrate the inner orientation elements and the distorted parameters based on the non-linear distortion model of camera. Experimental results of this method reveal good performance in the application of aerial triangulation operation and digital mapping.

Abstract:分析了Al2O3粒子在喷焰内流动冷却过程中不同相态对其光学性质的影响,重点研究凝固过程中三种相态(液态、γ相、α)Al2O3粒子的吸收指数随相态的变化关系,考虑了紫外到近红外光谱范围内,Al2O3粒子的不同辐射机制.基于Drude模型所得到的氧化铝粒子不同相态吸收指数的计算模型建立了描述Al2O3粒子相变过程紫外与可见光辐射特性数学计算模型.通过与实验数据的对比分析表明,本模型是可行的.在此基础上,分析了不同相态Al2O3粒子辐射特性随温度和波长的变化关系.

Abstract:介绍了待测样品在激光辐照下光热位移信号的理论表达式,阐述了信号相位与热扩散率以及调制频率之间的关系;基于该关系并结合相关资料,提出了一种新的采用迈克尔干涉检测装置测量热扩散率的简便方法,分析了其相对于传统测量方法的优势.采用该装置对Ti及TiO2复合薄膜样品进行了测量,由光热位移相位信号与调制频率之间的关系,计算得到了待测样品的热扩散率.

Abstract:针对同一场景红外图像与可见光图像的融合问题,提出了一种基于非采样Contourlet变换(Nonsubsampled Contourlet Transform,NSCT)图像融合算法.算法首先采用NSCT对源图像进行多尺度、多方向分解,得到低频子带系数和各带通方向子带系数.然后,针对低频子带系数的选择,提出了一种基于红外图像与可见光图像物理特征的"加权平均"系数选择方案;针对各带通方向子带系数的选择,结合人眼视觉特性,提出了一种基于区域能量匹配的系数选择方案,得到融合图像的NSCT系数.最后经过NSCT逆变换得到融合图像.实验结果表明该算法可获得较理想的融合图像,其融合效果优于传统的基于离散小波变换以及离散小波框架变换的图像融合算法.

Abstract: