Abstract:The influence of MS transport on the current voltage characteristic of MCT PV device was investigated based on the current voltage characteristics of MCT Schottkey barrier. The data of some devices were also discussed.

Abstract:利用共焦显微镜纵向扫描采样手段，发展了一种空间分辨（深度剖析）光谱方法进行荧光光谱和拉曼光谱的甄别以及它们相应跃迁的指认。利用荧光的自吸收现象，成功地甄别了Nd:YAG晶体中拉曼和荧光光谱，并详尽指认了荧光光谱线的相应跃迁归属。

Abstract:报道了用 MBE方法生长掺 In的 n型 Hg Cd Te材料的研究结果 .发现 In作为 n型施主在 Hg Cd Te中电学激活率接近 10 0 % ,其施主电离激活能至少小于 0 .6 m e V.确认了在制备红外焦平面探测器时有必要将掺杂浓度控制在～ 3× 10 1 5 cm- 3水平 .比较了高温退火前后 In在 Hg Cd Te中的扩散行为 ,得出在 40 0℃温度下 In的扩散系数约为10 - 1 4 cm2 / s,并确认了 In原子作为 Hg Cd Te材料的 n型掺杂剂的可用性和有效性 .

Abstract:采用飞秒脉冲光电探测技术研究低温生长砷化镓光电导开关超快瞬态响应特性.实验测得不同激发波长或偏置电压下LT-GaAs光电导开关瞬态响应驰豫时间约350～390fs,由实验数据计算得到电子迁移率约1000cm

Abstract:测量了几种不同处理的Cd1-xZnxTe(x=0.04)表面的傅里叶变换拉曼散射光谱和电流－电压（I－V）特性。通过分析拉曼光谱反Stokes分量，并与表面I－V特性进行比较，结果表明与表面处理相联系的晶格声子的行为反映了表面完整性的变化，Te沉淀是影响表面质量的关键因素，并对有关表面处理方法的实际应用进行了讨论。

Abstract:A new wavelet based image coding scheme was proposed. The Variable Coefficient Fixed Length(VCFL) Coding method was used to code the wavelet subband coefficients of the image. The codebook was obtained from symbol probability. By a way similar to Atithmetic Coding, a bunch of symbols was mapped to the code in the codebook. Several error concealement methods were also adopted to improve the robustness. Simulation results show that our new scheme has higher compression ratio and robustness.

Abstract:A new 3D image segmentation algorithm was proposed. This algorithm integrates the improved Active Contour model and a new forecast algorithm. The forecast algorithm is based on the peculiarity of 3D images i.e the deformation of the contour along the space or time axis is continuous in 3D images. The forecast algorithm analyzes the trend of the deformation of the contour in the segmented images, then prognosticates the location and the shape of the contour in the next image. Experiments on medical anatomic images show that the new algorithm can observably improve the accuracy of the segmentation and reduce the time needed.

Abstract:提出一种估计图像噪声的方法，该方法用混合高斯概率密度模型拟合图像的小波系数中最高频率子带的直方图，用EM算法估计模型的参数，选取其中最小的标准方差作为图像噪声标准方差。用该方法能准确地估计图像高斯噪声的标准方差，尤其当图像的噪声比较弱时，该方法比传统方法更准确。

Abstract:利用计算机对不同运动轨迹下ZnS头罩外表面膜厚分布进行了模拟，优化出头罩的最佳运动轨迹，在该轨轨迹下采用射频磁控反应溅射（RRFS）法进行头罩镀膜，能够得到满足使用要求的薄膜厚度均匀性。实验结果表明,头罩外表面薄膜厚度不均匀性小于10％，双面镀膜后，8－11.5 μm波段平均透过率从69.6%提高到87．2％以上，透过率的不均匀性小于1．2％，满足了红外应用中对ZnS头罩的要求。

Abstract:A novel pixel level image fusion scheme was presented based on multiscale decompositon. First, the wavelet transform is used to perform a multiscale decomposition of each image. Then, the wavelet coefficients of fused image are constructed using multiple operators according to different fusion rules. This approach is successfully used to fuse the infrared and visible light images. The experimental results show that the fusion scheme is effective and the fused images are more suitable for human visual or machine perception.

Abstract:The deficiency of original image transition region theory and its origin were analyzed. Based on the directional property of neighborhood pixels in the transition region, one dimensional linear regression process was introduced to eliminate the influence of random fluctuations in local image regions.The practical results show that the robustness and the accuracy of new algorithm could be improved greatly.

Abstract:提出了一种基于遗传算法的协同神经网络中参数的优化算法，该算法利用遗传算法的全局最优搜索能力，在协同神经网络的参数空间搜索最优解。对从“车牌识别系统”中采集得到的数字样本进行了测试表明：优化算法能有效提高协同神经网络的识别性能，使识别率达到了较为实用的水平（98．4％）。另外，还对协同神经网络中各个参数在识别过程中的作用进行了讨论。

Abstract:An algorithm for multisensor image fusion was proposed based on tree structure wavelet decomposition. Under some special energy rules, it can adaptively implement subband decomposition and image fusion according to the image features. Experimental results of visible image and infrared image fusion, visible image and millimeter wave image fusion were given, which show the proposed method can achieve better fusion performance than the traditional pyramid wavelet decomposition.

Abstract:A new wavelet kernel function network (WKFN) was proposed as an alternative of support vector machine (SVM). Simulation experiments show that the performance of approximation and recognition of WKFN is equal to that of the corresponding SVM, but gets an advantage over that of a general neural network.

Abstract:利用液相镀膜法研制成了性能较好的传输用聚苯乙烯－银－玻璃基体小直径空心波导，波导长度达到1m左右，波导直径为800－1200μm，对于CO2激光（10.6μm）的传输损耗低达1.73dB/m，损伤阈值高于11.2W(17.2W/mm^2)。实验结果表明，聚苯乙烯－银－玻璃基体小直径空心波导适用于医学等领域CO2激光传输的要求，并有望在其他红外波段的激光能量传输中得到应用。

Abstract:提出并实现了用微米级空间分辨率的显微光致发光（μ-PL)平面扫描谱对CdZnTe(CZT)晶片的表面亚微米层特性研究。在含缺陷区域进行微米尺度和在大面积范围内进行毫米尺度的逐点PL测量。对测得每一点的PL谱进行了拟合。拟合参数中等效温度Tc的统计分布给出两个分布中心，表明存在有两种机制的发光过程。同时统计结果给出发光各点的不均匀性。等效温度的平面分布图直观地给出了各温度的平面位置，样品经溴 抛光后重复类似的测量，结果表明等效湿度的统计均匀性大为改善。抛光后的不同的缺陷点表现出不同的发光特性，意味着各自起源的不同。大面积PL扫描的统计结果和平面分布给出样品特性的整体评价。

Abstract:In view of the characteristics of the low contrast and small dimension of targets and the complexity of background in infrared image, a method of detecting and describing infrared weak target was presented with improved K L transform and generalized K L transform. Thus the infrared weak target can be detected in low signal to noise ratio.