基于灰度跳变与字符间隔模式的车牌定位方法研究

车牌定位是车牌自动识别的第一步,而如何考虑光照影响是车牌定位是否成功的关键;通过深入分析不同的车牌图像,提出一种基于灰度跳变与字符间隔模式的车牌定位方法;首先,针对不同光照条件下采集到的车牌图像明暗度的不同,利用多阈值处理方法得到车牌信息不丢失的、最佳的二值图像,然后,在其二值图像中首先利用灰度跳变定位车牌的上下边界,接着对字符垂直投影后的宽度进行统一的调整并以固定的字符间隔特征定位车牌的左右边界,从而完成车牌定位;最后,通过实验验证了该方法的有效性。     

一种新型自适应环境光照的车牌定位算法

为解决在不同光照环境下车牌的适应性定位问题,同时解决基于颜色定位的方法普遍存在的车身颜色与车牌颜色相近,以至于难以区分的问题,分析了不同光照下的车牌图像在HSV空间的分布特性与图像亮度等参数关系,结合车牌的混色分布特点、RGB与HSV空间变量转换关系,提出了一种可直接对RGB空间进行处理、能适应光照环境的快速车牌定位算法;实验结果表明,提出的定位算法对光照不足、环境光不均匀等复杂条件下获取的蓝白车牌图像具有很高的定位可靠性。     

一种复杂车辆图像中的多车牌定位方法

针对复杂背景中多个车牌的定位问题,提出一种新的定位方法.该方法综合利用边缘检测、连通域分析、倾斜矫正等多种方法,解决了复杂背景中定位难的问题.能够准确定位杂乱背景中的车牌,对天气、光照变化、车牌在图像中的移动和旋转等,具有良好的适应能力.该方法为后续的字符分割和字符识别提供旋转角度、字符区域定位信息.     

基于OpenCV的车牌图像定位

现有的车牌定位方法几乎都需要先对白天夜晚的场景进行分类,在图像中存在其他灰度剧烈变化区域时,这种场景分类容易影响到车牌定位的准确率;为了对获取到的车辆图像进行准确地定位,提出了一种改进的灰度跳变车牌定位算法,首先加入了光照补偿,使得不需要对白天和晚上的场景进行分类;然后进行了长竖线的噪声去除,可以在背景比较复杂的情况下准确定位车牌;并对精确定位出的车牌进行边框的去除,有利于下一步的字符分割;另外,对OpenCV计算机视觉库的库函数的利用,算法的复杂度得到了简化,从而更好地满足了车牌识别系统的实时性;通过对出入口处采集到的分辨率为704×576的300张图片进行测试,定位率高于95%,对白天和夜晚获取的图像有很好的适应性,满足设计要求。     

基于STM32的车牌自动识别系统设计

介绍了一种基于STM32实现车牌识别系统,以STM32单片机作为控制器,控制OV7670图像传感器进行车牌图像信息获取,经过模式识别、二值化分析、识别车牌区域、字符定位、字符分割和字符识别,最后获取车牌字符,并在彩色液晶屏显示采集的图片和识别结果。设计结果证明该系统能把机动车车牌字符正确识别,且识别过程所用时间短,识别成功率高,具有一定的应用价值。     

车辆牌照识别系统的设计

利用数字图像处理技术以及神经网络技术对车牌识别系统进行研究。首先从采集的牌照图片中进行车牌定位,车牌字符分割,字符特征提取,最后送入BP神经网络进行识别。     

基于LabVIEW图像法车牌智能识别系统

车辆牌照自动识别是实现交通管理智能化的重要环节,设计中利用图像采集卡对经过的车辆车牌进行图像采集并传送至计算机,采用美国NI公司LabVIEW软件,实现图像预处理、图像去噪以及图像增强等功能;然后根据车牌颜色特征对其准确定位,采用阈值法分割车牌字符;最后由OCR函数来识别字符,识别结果保存至相应数据中,可以进行相应的违章、违规智能交通管理,经实验该系统成功实现车牌识别识别率达99%。     

基于SSIM的车牌识别算法研究

对车牌中的字符进行精确提取并识别是车牌识别系统面临的重要问题。在字符提取阶段,利用车牌图像特有的信息特征,提出以红色分量作为目标图像,依次进行水平和垂直投影完成字符分割,既可精确提取字符,又可矫正字符形态。将基于结构相似度SSIM(structure similarity)的算法应用于车牌识别,避免了前期对字符结构信息做大量的对比统计工作,省略了特征提取的步骤,降低了算法复杂度;利用角点检测方法对相似字符进行细化和区分,进一步完善了整体的识别性能。实验证明,该算法有较好的识别率。     

基于彩色像素统计与新型模板匹配的车牌识别系统

提出一种新型的车牌定位方法与字符识别方法。在整个车辆图片中,车牌区域具有对比度高,形状规整等特点,运用彩色像素统计法直接对车辆图像进行车牌定位,对提取的车牌进行滤波和锐化,抑制噪声对车牌区域的干扰,最后对类间距离差别较大的字符类型制作多个模板,来达到模板的更有效匹配。结果表明,与传统方法相比,该方法对车牌定位、字符的提取和识别具有精确度高、运算时间少且鲁棒性好的优势。     

快速定位车牌的方法研究

提出了基于数学形态学和纹理特征的快速定位车牌的方法.该方法包含4个步骤:预处理,利用Sobel算子进行垂直边缘检测,通过形态学操作得到连通域,利用车牌的纹理特征和字符特征最终定位车牌.使用该方法对1 144幅图片进行实测,1 102幅图片定位成功,正确识别率为96.3%,处理时间为160～200ms.     

未系安全带智能识别中安全带定位关键方法研究

安全带定位是实现机动车未系安全带智能识别的关键。针对道路监控图像特点,提出一种基于梯度变换的安全带定位方法。该方法在对卡口图像进行预处理的基础上,采用自适应阈值边缘检测算法以及积分投影方式定位车辆位置,设立车牌检测区域,以减少运算量,降低干扰,同时利用训练得到的Haar分类器识别车牌位置,通过逼近方式切取车辆右侧图像。最后采用梯度变换算法求得车窗各边缘坐标,实现车窗精准定位,并计算得到安全带位置。试验表明,该方法可实现安全带的准确定位,具有较好的实用性,为后续安全带识别奠定基础。     

License plate location based on trichromatic imaging and color-discrete characteristic

License plate location (LPL) is the key part of automatic license plate recognition (LPR) system that plays an important role in many applications and a number of LPL methods based on color information have been proposed. However, some problems, such as country specific, similar color interference, lighting variation sensitive, time consuming, processing only one kind of LP with same color combination a time and omitting the case of more than one combinations in a LP, should be solved further. In this study, a color-based LPL method that consists of three modules: color edge extraction, denoising and searching, is proposed. Color edge extraction, the kernel of the proposed, is designed by color-discrete characteristics of license plates in the trichromatic wavebands. In the experiment, 1384 images taken from natural scenes in China and other 104 countries or regions are employed. Of which, 74 have been failed to locate the license plates. The success rate and average execution time are 94.7% and 57 ms, respectively.     

License plate detection using gradient information and cascade detectors

In this paper, we propose a license plate detection method based on the gradient information and a cascade detection framework. The proposed method consists of three main modules: image preprocessing, license plate detection and license plate confirmation. Considering the diversities of Chinese license plate, we first get the gradient images to unify plate forms through image preprocessing. The license plate will be detected roughly by using the cascade AdaBoost classifier in the next step. Finally, several heuristic judgment strategies and a voting-based method are used to verify the candidate license plates. The effectiveness of our method is validated with an experimental study in license plate detection.     

Analytical modeling and vibration analysis of internally cracked rectangular plates

This study proposes an analytical model for nonlinear vibrations in a cracked rectangular isotropic plate containing a single and two perpendicular internal cracks located at the center of the plate. The two cracks are in the form of continuous line with each parallel to one of the edges of the plate. The equation of motion for isotropic cracked plate, based on classical plate theory is modified to accommodate the effect of internal cracks using the Line Spring Model. Berger?s formulation for in-plane forces makes the model nonlinear. Galerkin?s method used with three different boundary conditions transforms the equation into time dependent modal functions. The natural frequencies of the cracked plate are calculated for various crack lengths in case of a single crack and for various crack length ratio for the two cracks. The effect of the location of the part through crack(s) along the thickness of the plate on natural frequencies is studied considering appropriate crack compliance coefficients. It is thus deduced that the natural frequencies are maximally affected when the crack(s) are internal crack(s) symmetric about the mid-plane of the plate and are minimally affected when the crack(s) are surface crack(s), for all the three boundary conditions considered. It is also shown that crack parallel to the longer side of the plate affect the vibration characteristics more as compared to crack parallel to the shorter side. Further the application of method of multiple scales gives the nonlinear amplitudes for different aspect ratios of the cracked plate. The analytical results obtained for surface crack(s) are also assessed with FEM results. The FEM formulation is carried out in ANSYS.     

基于多核的车牌识别的架构实现

多核技术是现在提高芯片性能的主要方法。区别于传统以PC和DSP为核心的车牌识别系统,以FPGA为核心,利用SOPC技术构建了车牌识别多核处理器。给出了一种基于多核的车牌识别架构,在该多核处理器中,以3个Nios II 软核为主要处理器核处理车牌定位、字符特征识别提取及识别等处理,同时构建硬件加速器作为协处理器处理图像增强、边缘检测和膨胀、腐蚀等数学形态学处理。在CQ片上路由器基础上,构建了NOC用以实现片上多核通信。另外,为了保证路由器与多处理器核之间的快速、并行通信,加入了数据驱动模块。整个系统在Altera Cyclone IV FPGA上实现了车牌的识别。这种片上系统设计方法具有硬件设计灵活,可扩展性强等优点,能有效地降低系统软硬件设计的难度,缩短开发周期,并提高设计的可靠性。     

Vibration of L-shaped plates under a deterministic force or moment excitation: a case of statistical energy analysis application

Analytical and closed form solutions are presented in this paper for the vibration response of an L-shaped plate under a point force or a moment excitation. Inter-relationships between wave components of the source and the receiving plates are clearly defined. Explicit expressions are given for the quadratic quantities such as input power, energy flow and kinetic energy distributions of the L-shaped plate. Applications of statistical energy analysis (SEA) formulation in the prediction of the vibration response of finite coupled plate structures under a single deterministic forcing are examined and quantified. It is found that the SEA method can be employed to predict the frequency averaged vibration response and energy flow of coupled plate structures under a deterministic force or moment excitation when the structural system satisfies the following conditions: (1) the coupling loss factors of the coupled subsystems are known; (2) the source location is more than a quarter of the plate bending wavelength away from the source plate edges in the point force excitation case, or is more than a quarter wavelength away from the pair of source plate edges perpendicular to the moment axis in the moment excitation case due to the directional characteristic of moment excitations. SEA overestimates the response of the L-shaped plate when the source location is less than a quarter bending wavelength away from the respective plate edges owing to wave coherence effect at the plate boundary.     

Self-adaptive model of texture-based target location for intelligent transportation system applications

In this paper, three shortages: full-enhancement mechanism, invariable texture extraction mechanism, and lack of robust enhancement methods, that are always neglected in some tasks based on texture-based target location for Intelligent Transportation System applications (ITS) are introduced. To cover these shortages, a self-adaptive model for texture-based target location is established. It is mainly composed of self-adaptive enhancement mechanism, self-adaptive texture extraction mechanism, and graying enhancement. Also, feedback evaluation strategy for design of texture extraction modes, needed by self-adaptive texture extraction mechanism, is provided. And then, an algorithm of license plate location is designed to evaluate effectiveness of this model. 338 license plate images from natural scenes were applied to test graying enhancement and self-adaptive location. Results of enhancement experiment indicate that graying enhancement is robust in purposiveness, real-time performance, and empirical parameter independence, simultaneously. Further, Results of location experiment demonstrate that not only a sound algorithm of license plate location with high success rate (96.1%) and low average execution time (52 ms) is given but also the effectiveness of the two self-adaptive mechanisms in the proposed model are validated.     

基于BP神经网络的车牌字符识别算法研究

车牌字符识别是车牌识别系统中的关键环节。采用图像处理和神经网络相结合的方法设计新的车牌字符识别算法,先对分割出的车牌字符进行归一化处理,然后进行SOBEI．边缘检测和角点特征提取,最后输入BP神经网络进行训练、识别,其中BP神经网络模型属于改进型神经网络。通过一系列神经网络训练和仿真实验,车牌识别速度和正确率得到了明显的提高。     

A note on the damping of large amplitude beam vibrations

This paper presents a new series-type method for solving the eigenvalue problems of irregularly shaped plates clamped at all edges. An irregularly shaped plate is formed on a simply supported rectangular plate by rigidly fixing several segments. With the reaction forces and moments acting on all edges of an actual plate of irregular shape regarded as unknown harmonic loads, the stationary response of the plate to these loads is expressed by the use of the Green function. The force and moment distributions along the edges are expanded into Fourier series with unknown coefficients, and the homogeneous equations for the coefficients are derived by restraint conditions on the edges. The natural frequencies and the mode shapes of the actual plate are determined by calculating the eigenvalues and eigenvectors of the equations. The method is applied to a cross-shaped, an I-shaped and an L-shaped plate clamped at all edges, the natural frequencies and the mode shapes of the plates are calculated numerically and the effect of the shape is discussed.