自然场景图像的字符识别方法

目的基于大津算法(Otsu算法)对图像进行分割,利用光学字符识别方法对自然场景图像中的英文字符进行识别。方法首先用分块Otsu算法对图像进行初步的二值化,然后通过对二值化结果的分析,把原始的输入图片分割成单个字符的子图,再对各子图重新用Otsu算法进行二值化,最后对最终得到的二值化结果进行识别,再结合之前得到的每幅图的字符数量信息和词典信息,对识别结果进行修正,得到最终的识别结果。结果在ICDAR2013数据集上测试文中算法,单词正确识别率为46.03%,总编辑距离为474.5。结论文中提出的以Otsu为基础的分块识别算法,能够更好地分割复杂背景图像的背景和文本,同时结合词典信息对识别结果进行了修正,改善了识别效果。     

基于图像奇异值描述的雷达目标识别方法

利用目标对常规雷达发射波形的调制效应，先对录取的回波数据进行预处理，把一维回波序列投影成二维灰度图像，然后提取这种图像的反映飞机目标不同机型(大、小)和不同架次的奇异值作为特征矢量，采用BP神经网络对目标进行分类识别试验，结果表明该方法是有效的，这为常规低分辨雷达空中目标识别提供了一种新方法。     

字符形态特征的机械仪表判读算法

基于字符形态特征提出一种快速的机械仪表判读算法.该方法中采用平滑滤波对灰度化后的图像去除噪音,再使用OTSU算法进行二值化,同时根据图像投影确定表盘区域,最后将Canny算子和Hough变换结合对图像进行矫正并分割字符.字符判读模块中对图像进一步去除噪音,对字符形态特征进行加权计算和对比,准确识别出仪表示值.通过实验证明,该算法辨识率高,有较好的实用性.     

基于形态学Top-Hat算子实现运动图像的分割与识别

针对高速运动图像的目标识别问题,首先运用数学形态学方法进行预处理,然后采用阈值分割对待识别图像进行二值化处理,最后,对分割后的图像进行链码跟踪,利用Hu不变矩识别运动目标。试验结果表明,通过设计特殊的结构元素,利用数学形态学的Top-Hat算子,可以对高速运动图像进行有效分隔,利用Hu不变矩能够快速准确的识别空间运动目标。     

基于二值随机森林的非均匀光照QR码重构算法

目的 现实生活QR码在识别过程中,易受到非均匀光照因素的影响,导致QR码无法识别,为此提出一种基于二值随机森林的QR码像素值重构算法。方法 依据QR码图像的双峰特点和梯度值相等的特性,用于提取非均匀光照下受损QR码局部像素特征,并利用随机森林的分类方法确定QR码局部矩阵中间单个像素值,逐步实现受损QR码所有像素值的重构恢复。结果 实验表明与其他方法相比,该算法模型能够很好地利用局部特征提取QR码的真实像素值,并对受损QR码图像进行恢复,实验结果图像均具有较高水平。结论 采用基于二值随机森林的QR码重构算法,能够很好地处理因非均匀光照而导致的识别出错问题,并可以广泛应用于生活中的QR码识别过程,具有较强的实用性。     

基于空间邻域信息的二维模糊聚类图像分割

传统模糊C均值聚类(FCM)算法进行图像分割时仅利用了像素的灰度信息,并且使用对噪声较敏感的欧氏距离作为像素与聚类中心距离度量的标准,因此抗噪性能较差.为了克服传统FCM算法的局限性,本文提出了一种基于空间邻域信息的二维模糊聚类图像分割方法(2DFCM).该方法利用二维直方图描述的像素邻域关系属性,一方面为聚类提供较准确的初始聚类中心,从而避免聚类中的死点问题;另一方面通过提出聚类中心同时在像素值、像素邻域值二维方向上进行更新的思想,建立了包含邻域信息的新的聚类目标函数,实现了图像的分割.实验结果表明,这种方法抗噪能力强、收敛速度快,是一种有效的模糊聚类图像分割方法.     

基于类圆分割的棒材计数图像识别

棒材集捆后的自动计数是个一直没有得到很好解决的问题，针对棒材的特点，提出了一种基于类圆形假设的图像分割、识别方法。该方法对多粘连的棒材端面图像可以进行有效的分割，提高了物体识别和计数的准确率。同时，由于算法中基本上都是加法、减法和逻辑运算，速度也得到了改善。在对实物图像的实验中，新算法得到了较好的结果，尤其是在对多根螺纹钢粘连的图像的处理上取得了成功。     

局部高亮干扰文本图像的二值化方法研究

本文提出一种新的基于Curvelet变换的文本图像二值化处理方法,以消除文本图像中局部高亮度区域对二值化图像质量的影响.首先对具有局部高亮度区域干扰的原始文本图像进行Curvelet变换,得到图像在曲波域的Curvelet系数集；然后根据各Curvelet系数所表征的图像特征,对Curvelet系数进行非线性增强,以优化文本图像的直方图分布；对增强的Curvelet系数集进行反变换,得到直方图优化后的时域图像,进而应用Otsu方法实现文本图像二值化.应用本文方法对具有带状及点状局部高亮度区域的文本图像进行二值化处理,并采用ABBYYFineReaderl0对二值图像进行OCR识别.实验结果表明,通过本文提出的处理方法所得到的二值化图像,其字符的OCR识别准确率最高可达94.81％,优于其他四种典型的图像二值化处理方法.     

模糊理论与BP网络在目标识别中的应用

针对利用神经网络进行目标识别时特征向量选取中存在的一些问题：如特征向量选取不当，导致不同目标特征向量值可区分性差；相同目标由于大小、平移、旋转角度的不同，导致特征向量值具有较大差异等，首先对样本图像边缘提取，然后对已有的隶属函数进行改造，提出了一种基于模糊理论的阈值分割法，把图像二值化处理，提取出样本图像中目标的边缘轮廓，对其取不变矩．并归一化不变矩．为了避免不变矩数值过小，对其取对数，以此作为BP网络的输入特征向量，进行训练和识别．试验表明该方法能快速有效地识别出目标．     

基于平滑卷积定位的圆形印鉴提取方法研究

利用圆形印鉴中心的五角星所在区域红色像素分布集中的特点,提出了一种基于平滑卷积定位的圆形印鉴提取算法.该算法首先将彩色票据图像转换为灰度图像,接着利用平滑卷积定位五角星区域中的一个像素点,以这个像素点邻域的平均灰度值作为阈值二值化票据图像,然后搜索圆形印鉴边框,计算圆心坐标和半径,从而实现印鉴的提取.同时,为了提高提取速度,通过适当减小卷积过程中图像的分辨率,优化了整个提取过程中时间复杂度最高的平滑卷积算法.实验结果表明,这种方法能够在复杂背景下有效地提取出圆形印鉴,在经过算法优化后,提取速度提高了近12倍.     

低信噪比图象识别中二元统计编码的应用

用数理统计观点，把图象看成是由纯噪声象和纯目标象叠加而成，用二元统计编码方法，对信噪比接近１的相干图象，进行了计算机模拟，获得了被噪声淹没的二元目标象，给出了对比度与目标恢复程度之间的对应关系。     

Two-echelon spare parts inventory system subject to a service constraint

In this paper, we consider a spare parts inventory problem faced by a manufacturer of electronic machines with expensive parts that are located at various customer locations. The parts fail infrequently according to a Poisson process. To serve customers when a failure occurs, the manufacturer operates a central warehouse and many field depots that stock spare parts. The central warehouse acts as a repair facility and replenishes stock at the field depots. There is a centralized decision-maker who manages the inventory in both the central warehouse and the field depots. We develop a continuous review, base stock policy for this two-echelon, multi-item spare parts inventory system. We formulate a model to minimize the system-wide inventory cost subject to a response time constraint at each field depot. We present an efficient heuristic algorithm and study its computational effectiveness.     

Remote Sensing Image Classification Algorithm Based on Texture Feature and Extreme Learning Machine

With the development of satellite technology, the satellite imagery of the earth’s surface and the whole surface makes it possible to survey surface resources and master the dynamic changes of the earth with high efficiency and low consumption. As an important tool for satellite remote sensing image processing, remote sensing image classification has become a hot topic. According to the natural texture characteristics of remote sensing images, this paper combines different texture features with the Extreme Learning Machine, and proposes a new remote sensing image classification algorithm. The experimental tests are carried out through the standard test dataset SAT-4 and SAT-6. Our results show that the proposed method is a simpler and more efficient remote sensing image classification algorithm. It also achieves 99.434% recognition accuracy on SAT-4, which is 1.5% higher than the 97.95% accuracy achieved by DeepSat. At the same time, the recognition accuracy of SAT-6 reaches 99.5728%, which is 5.6% higher than DeepSat’s 93.9%.     

SVD-based digital image watermarking using adaptive generated watermark

In this paper, an evolutionary-based algorithm for digital image watermarking is proposed. Robustness and imperceptibility are two important properties of digital image watermarking which conflict with each other. A singular value decomposition (SVD)-based intelligent algorithm is applied to make a trade-off between robustness and imperceptibility. In the proposed algorithm, first, the original image is divided into 8×8 blocks. Then some special blocks are transformed into discrete cosine transform (DCT) domain. SVD decomposition of DCT coefficients of special blocks is the next stage of the method. Finally, a binary watermark image is embedded into the singular values by a quantisation method. The major property of the proposed method is the generation of a binary watermark image using genetic algorithm to solve the optimisation problem between robustness and imperceptibility. So the watermark could be variable and image adaptive. The simulation results show improvement in robustness of the proposed method against several attacks such as noise contamination, filtering, joint photographic experts group (JPEG) compression and some geometrical attacks like cropping, rotation and rescaling in comparison with a recently similar existing work.     

利用粗糙集和属性直方图的图像增强方法

利用粗糙集理论进行图像增强,子图的划分是关键。属性直方图是对直方图概念的推广,是一种由先验知识约束的直方图;将它用于子图的划分,在此基础上提出了一种基于粗糙集理论和属性直方图的图像增强方法。该方法利用属性直方图的 Otsu 算法确定灰度阈值,根据灰度阈值利用不可分辨关系,将图像划分为背景子图、目标子图和噪声子图,对去噪后背景子图和目标子图进行增强变换,并将它们合并得到增强图像。将该方法用于一种海底小目标图像增强。实验结果表明该方法处理增益为 11dB,明显地增强了图像,且不损害图像的边缘。该方法适用于图像有某种先验知识的场合。     

结合小波与增量符号相关的鲁棒图像匹配

结合小波变换和增量符号相关提出了一种有效的鲁棒图像匹配方法。将实时图和基准图做小波分解以获得粗尺度的平滑图像;利用相邻像素间灰度的升降对所获得的平滑图像进行二值化重新编码;匹配二值化后的实时图和基准图,找出前者在后者中的坐标,进而确定实时图在基准图中的位置,以完成匹配。小波分解能有效的去除噪声并能大大提高匹配算法的实时性。增量符号相关方法对图像灰度值的变化不敏感。因此,二者相结合不仅有效的解决了存在噪声干扰、阴影混入（块误差）和部分遮挡等影响下的图像匹配问题,且与传统匹配方法相比,匹配速度提高了约10倍以上。理论分析说明了方法的可行性,多组实验显示了其有效性。     

Simultaneous measurements of skin sea surface temperature and sea surface emissivity from a single thermal imagery

A novel method, to our knowledge, to measure simultaneously the thermal emissivity and skin temperature of a sea surface has been developed. The proposed method uses an infrared image that includes a sea surface and a reference object located near the surface. By combining this image with sky radiation temperature, we retrieve both skin sea surface temperature and sea surface emissivity from the single infrared image. Because the method requires no knowledge of thermal radiative properties of actual sea surfaces, it can be used even for a contaminated sea surface whose emissivity is hard to determine theoretically, e.g., oil slicks or slicks produced by biological wastes. Experimental results demonstrate that the estimated emissivity agrees with the theoretical prediction and, also, the recovered temperature distribution of skin sea surface has no appreciable high-temperature area that is due to reflection of the reference object. The method allows the acquisition of match-up data of radiometric sea surface temperatures that precisely correspond to the satellite observable data.     

Solving large multiple-depot multiple-vehicle-type bus scheduling problems in practice

We consider the multiple-depot multiple-vehicle-type scheduling problem (MDVSP) which arises in public transport bus companies and aims to assign buses to cover a given set of timetabled trips with consideration of practical requirements, such as multiple depots and vehicle types as well as depot capacities. An optimal schedule is characterized by minimal fleet size and minimal operational costs including costs for empty movements and waiting time. It is well-known that the MDVSP is NP-hard.Although progress has recently been made in solving large practical MDVSP to optimality with time-space network models, current optimization technology sets limits to the model size that can be solved. In order to approach very large practical instances we propose a two-phase method which produces close to optimal solutions. This modeling approach enables us to solve real-world problem instances with thousands of scheduled trips by direct application of standard optimization software. Furthermore, we introduce the concept of depot groups for the case that a bus may return in the evening into another depot than where it started in the morning.     

大孔径静态干涉成像光谱仪图像校正技术

设计了一种基于相位相关与归一化积相关的联合图像校正算法。该算法针对卫星姿态变化对LASIS(大孔径静态干涉成像光谱仪)图像造成的畸变,根据傅里叶变换的旋转平移特性,实现图像序列旋转失真的高精度校正,校正精度达到0.01;同时,基于归一化积相关方法实现图像序列平移失真的校正。针对LASIS数据量大的特点提出了改进的快速自适应模板选取法,使算法的运算复杂度按4n递减。仿真结果表明,该算法能快速有效地校正LASIS图像。     

基于双正交小波变换的资源二号卫星与多光谱影像融合方法研究

针对基于Darbechies等正交小波函数的遥感影像融合时引起失真的现象,引入双正交小波函数,提出了基于双正交小波变换的融合方法,对资源二号卫星和TM的多光谱影像的融合进行研究,并把融合结果与主成分、IHS和Brovey等融合结果进行比较。通过对光谱特征曲线、相关系数、光谱扭曲程度等分析和空间细节的对比,说明双正交小波融合方法除了提高空间分辨率外,最好地保持了多光谱影像的光谱特征,有利于专题信息的提取。