Two-tier architecture for unconstrained handwritten character recognition

In this paper, we propose an approach that combines the unsupervised and supervised learning techniques for unconstrained handwritten numeral recognition. This approach uses the Kohonen self-organizing neural network for data classification in the first stage and the learning vector quantization (LVQ) model in the second stage to improve classification accuracy. The combined architecture performs better than the Kohonen self-organizing map alone. In the proposed approach, the collection of centroids at different phases of training plays a vital role in the performance of the recognition system. Four experiments have been conducted and experimental results show that the collection of centroids in the middle of the training gives high performance in terms of speed and accuracy. The systems developed also resolve the confusion between handwritten numerals.     

流动单元研究在油田开发中的应用

流动单元研究提供了精细解剖砂体，表征砂体内部非均质性的方法。应用流动单元研究的模式及不同层次流动单元的研究方法，对油田开发后期以结构要素为基本单位的剩余油研究和井网高速思路等方面进行了探讨。以东部某油田E1f1层为研究对象，对其储层流动单元分布进行了划分。应用途径包括：指导油水井部署，对高产部位进行预测，定量分析储量动用状况等。     

一个基于相关存储的模式识别方法

应用相关的存储技术实现模式识别可缩减存储大小,但由此带来的饱和问题严重影响系统正常工作,为解决此问题,本文在分析现有系统的基础上给出了基于一系列逻辑操作的识别技术.该技术的实质是用一形式化的方法增加输入模式间的线性无关性.通过论证与实例验证,该方法是有效而实用的.     

检测骨架图形特征点的新方法

端点、折点、歧点和交点是骨架图形的重要特点是，提出用快速查表法检测侯选特征点，再进行侯选特征点聚集，用局部结构分析，去除假特征点，判别真特征点的类型，并对折点给以大致的定量描述。实验结果表明，该方法能快速有效地找出骨架图形的特征点，在目标识别、字符识别等应用中起着关键作用。     

用于一维图像识别的支撑矢量机方法

研究了支撑矢量机的分类机理,并利用支撑矢量机对雷达目标一维像进行了识别.识别的结果表明了该方法的优越性,并显示它可以对残缺不全的样本进行识别.     

基于距离信息扩展原理的物体识别方法

利用稀疏距离图象与灰度图象相结合，通过射影变换可使灰度图象特征有条件地获取距离信息，这称为 基于射影变换的距离信息扩展原理。利用这些距离信息可计算物体表面平面图形的欧氏空间不变量并对景物中的 物体产生假设。由于初始假设建立在可靠的基础上，并使用了带反馈的混合控制策略，使这种基于模型的物体识别 定位技术具有实用的前景。此文讨论了整个方法的原理及其实施方案，并实验验证了原理的正确性与有效性。     

Hiding iris data for authentication of digital images using wavelet theory

This paper introduces an efficient approach to protect the ownership by hiding an iris data into digital image for an authentication purpose. It is based on the theory of wavelets. The idea is to secretly embed biometric data (iris print) in the content of the image identifying the owner. The system is based on an empirical analysis of biometric and watermarking technologies, and it is split into several processes. The first process is based on iris image analysis, which aids the generation of the iris code (watermark); the second and the third processes deal with embedding and detecting a watermark; and the last process deals with the authentication. A new metric that measures the objective quality of the image based on the detected watermark bit is introduced, which does not require the original unmarked image for watermark detection. Simulation results show the effectiveness and efficiency of the proposed approach. The text was submitted by the author in English.     

全双极WTA模型的改进及其实现

提出了一种相似度的新定义，使双极ＷＡＴ模型的实现更加简洁明了。增加一个阈值使系统能判别非存储模式或因其信息量过低而无法鉴定的输入模式，阈值的大小与要求的容错能力有关。双极相似度及阈值通过光学方法实现。阈值的光学实现使系统仍具有照明不变性。通过电路实现取及ＷＴＡ网络的迭代运算。     

LiF晶体中F+,/sup>2色心在室温下的激光运转及可调谐输出

本文把纯LiF晶体中色心激光宽带输出的转换效率提高到11%，得到了心激光的宽带输出与温度的变化曲线。实现了色心激光的波长从0.86～1.02 μm的可调谐输出，并对泵浦光作用下LiF晶体中的各种色心动力学机制进行了分析。     

车号定位识别与可信度估计

以车号的定位与识别为例，应用行投影法对车号图片进行倾斜行定位，证明了模拟与场景的二维互相关系数近拟等于一维投影互相关系数，在此基础上提列投影法快速行扫描识别算法及相应的可信度估计方法，这种方法对于室外拍摄的有复杂背景，光照变化，字体微倾，字符排列疏密不均及噪声干扰的车牌图像具有很好的识别效果。