图像奇异值特征矢量缩放不变性分析及应用

图像奇异值特征矢量以其稳定性和它在转置、平移、旋转和镜像变换下具有的不变性,被认为是一种图像的代数特征,广泛应用于图像匹配和图像识别。通过矩阵运算,证明了奇异值特征矢量在图像做缩放变换时具有不变性,并将这种性质运用于图像匹配。实验结果表明这种不变性与奇异值特征矢量在图像做转置、平移、旋转和镜像变换时所具有的不变性一样,是奇异值特征矢量能够作为图像代数特征的一个重要依据。     

Robust图像跟踪：强度—特征混合式目标定位

在像平面上确定目标的位置(由它的瞄准点位置表示)是图像跟踪技术的核心.目前研究的目标是图像定位方法多偏重于单一的工作模式,因而难以适应变化较大的使用环境.本文提出一种旨在综合利用强度相关和特征匹配的混合式目标图像定位方法.假定当前帧图像已经过预处理并完成分割,首先通过一种自适应门限SSDA并辅之以投影排序搜索,尽可能快速地剔除与参考图差别较大的子区;再用一种简便可靠并可具一定记忆能力的特征匹配法,在保留的子区集内,为目标精确定位——宣布瞄准点位置.利用实际序列图像所作实验表明:上述方法较之通常采用的单一工作模式的定位方法,有明显增强的Robust性以及更广的适应性,而增加的计算代价不大.     

基于轮廓线匹配的二维重迭物体的识别

本文提出一种基于轮廓线匹配的二维重迭物体的识别算法,并采用一种点模式匹配的快 速算法进行特征点序列的匹配,通过引进形状移位数的概念和利用问题中的各种约束条件,提 高了物体识别速度.     

Ku波段10W功率PHEMT

报告了研制的 9.6mm栅宽双δ-掺杂功率 PHEMT,在 fo=1 1 .2 GHz、Vds=8.5 V时该器件输出功率3 7.2 8d Bm,功率增益 9.5 d B,功率附加效率 44.7% ,在 Vds=5～ 9V的范围内 ,该器件的功率附加效率均大于42 % ,两芯片合成 ,在 1 0 .5～ 1 1 .3 GHz范围内 ,输出功率大于 3 9.92 d Bm,最大功率达到 40 .3 7d Bm,功率增益大于 9.9d B,典型的功率附加效率 40 %。     

钨金属化与氧化铝陶瓷高温共烧

阐述了钨金属化与氧化铝陶瓷高温共烧工艺的特点,分析了烧结过程中温度曲线和露点等工艺参数对钨金属化与瓷件结合强度、密封气密性和收缩率的影响。通过反复的试验和烧结机理分析对比,给出了最佳的控制参数。     

2.6～2.8GHz 100W硅微波脉冲功率晶体管

研究并制造了2.6～2.8GHz输出功率100W、脉宽1μs、占空比25%、增益7dB、效率35%的硅微波脉冲功率晶体管。本文介绍了该器件的设计考虑和采用的工艺技术,给出了研制结果。     

HPD-2型非临氢降凝催化剂工艺实验

利用HPD-2型非临氢降凝催化剂进行了中试实验。结果表明：HPD-2型非临氢降凝催化剂相对其他的同类催化剂产品液体收率较高，最高可达95％以上；在反应温度为380℃时，凝点降幅在50℃左右，降凝效果较好；产品性质稳定，具有很好的应用前景。     

PVT Measurements on tetrafluoroethane (R134a) along the vapor-liquid equilibrium boundary between 288 and 373 K and in the liquid state from the triple point to 265 K

For the investigations of the gas-liquid phase equilibria, a new apparatus has been developed capable of simultaneously determining the pressure and the liquid and vapor densities using Archimedes' principle. The relative measurement uncertainties of the liquid and vapor densities of R134a (purity, 99.999%) at 313 K are 2×10^–4 and 7×10^–4, respectively (95% confidence level). For the measurements in the liquid region along nine quasi-isochores at pressures up to 5 MPa, an isochoric apparatus was used. The relative measurement uncertainty ofpv/(RT) is less than 1×10^–3. In addition to the investigation of the (p, v, T) properties, the temperature and pressure at the triple point and the vapor pressure between the triple point and 265 K were measured. On the basis of these data, a vapor pressure correlation has been developed that reproduces the measured vapor pressures within the uncertainty of measurement. The results of our measurements are compared with a fundamental equation for R134a, which is based on the measurements of other research groups.Paper presented at the Twelfth Symposium on Thermophysical Properties, June 19–24, 1994, Boulder, Colorado, U.S.A.     

Rheology in Processing of in Situ Composites of Polypropylene and Low Melting Poing Metals

The low melting point metallic tin powder or alloy of tin and lead was blended with polypropy-lene. A kind of in situ composite has been prepared. The variations of torque were studied when the composites were mixed in Haake torque rheogeniometer. By way of capillary extrusion, effects upon rheology of the in situ com-posites of the low melting point metals (LMPM) and coupling agent for their different variety and content, were investigated. From flow curves, the results indicate that in situ composites mixed with the LMPM are a kind of pseudoplastic fluid. If the LMPM were melted, the higher the content of the LMPM , the lower apparent viscosity of composites. Meanwhile, when the coupling agent is added into composites, the viscosity of composite will go up first and drop then. This shows that the LMPM have a promoter flow action on the polypropylene.     

A Multi-View Face Recognition System

In many automatic face recognition systems,posture constraining is a key factor preventing them from application.In this paper a series of strategies will be described to achieve a system which enables face recognition under varying pose.These approaches include the multi-view face modeling,the threschold image based face feature detection,the affine transformation based face posture normalization and the template matching based face identification.Combining all of these strategies,a face recognition system with the pose invariance is designed successfully,Using a 75MHZ Pentium PC and with a database of 75 individuals,15 images for each person,and 225 test images with various postures,a very good recognition rate of 96.89% is obtained.