共查询到17条相似文献,搜索用时 312 毫秒
1.
2.
3.
车辆牌照的自动识别是智能交通系统中的一项重要技术,而车辆牌照的定位又是车牌识别的关键点之一。文章依据二值化图像中车牌区域跳变频率高的事实,提出一种算法来确定车辆牌照在原始图像中的水平位置和垂直位置,从而定位车辆牌照。实验结果表明本算法处理速度较快、便于实现。 相似文献
4.
针对序列图像的多车牌定位问题,提出了一种综合车牌特征与梯度分析的定位算法.该算法使用灰度均衡化与中值滤波算法对图像预处理来消除噪声,利用测试样本图像提取出若干有效的车牌参教作为定位系统的输入,实现了全自动定位车牌的目的.并引入MGD梯度分析方法辅助边缘提取算法来提取车牌候选区域,利用HSV颜色模型,使用主颜色分析方法分析候选区域,并通过车牌字符纹理分析方法做进一步的判断.使用Hough轮廓检测方法计算非文本区域块的直线斜率.该算法定位效果较好,鲁棒性强,有很好的工程应用推广价值. 相似文献
5.
车牌识别中反色判断的新方法 总被引:3,自引:0,他引:3
根据车牌的基本特征,提出一种将车牌候选区域的灰度图颜色统一转换成白底黑字(或黑底白字)的新方法。该方法通过计算车牌灰度图像灰度平均值和大于灰度平均值的像素数,来判断车牌是否需要进行反色处理。实验表明,该方法简单易行,且准确率较高。 相似文献
6.
7.
8.
9.
变化光照条件下交通标志检测算法的准确率往往会显著降低。针对此问题,提出了1种新颖的概率图建立方法,并结合最大稳定极值区域特征进行交通标志检测。该方法包括3个处理步骤:①根据不同光照条件对真实场景交通标志样本图像进行明确分类以构建多类颜色直方图,将交通标志输入图像由原始色彩表达转变为概率图(直方图反投影);②通过在概率图上进行 MSER特征提取,获取候选的交通标志区域;③根据候选区域的面积、宽高比等特征快速有效去除非交通标志区域。实验结果表明在弱光照和强光照条件下基于归一化RGB的交通标志检测算法检测准确率分别下降到84.4%和83.0%,基于红蓝图的交通标志检测算法检测准确率分别下降到87.4%和86.3%,提出的算法在变化光照条件下依然可以保持90%以上的检测准确率,对光照变化有较好的鲁棒性。 相似文献
10.
基于彩色图像车牌分割研究 总被引:4,自引:1,他引:4
智能运输系统中车牌识别技术得到了广泛应用,车牌分割是车牌识别的重要部分。基于彩色图像车牌分割与采用灰度图像车牌分割相比,可以有效消除阴影影响,同时车牌颜色也是车牌识别的一个参数。颜色分类处理使用特征函数,可以减少颜色坐标转换运算,提高颜色分类速度。文中详细讨论中国车牌特征,给出车牌分割详细步骤。车牌区域判别采用信息融合技术。车牌倾斜矫正结合车牌倾斜特点,提出快速算法。 相似文献
11.
License plate extraction method for identification of vehicle violations at a railway level crossing
B. K. Cho S. H. Ryu D. R. Shin J. I. Jung 《International Journal of Automotive Technology》2011,12(2):281-289
The primary cause of most railroad accidents is vehicle entry into railway level crossings despite warning messages. To identify
drivers who violate railway level crossing regulations, vehicle license plate recognition can be applied at railway level
crossings. The purpose of this paper is to present an effective method for extracting the license plate region from vehicle
images taken at railway level crossings. The method proposed in this paper uses the variation in the gray-level values across
the image of a license plate. For license plate region extraction, the character region is first recognized by identifying
the character width and the difference between the background region and the character region. The license plate region is
then extracted by finding the inter-character distance in the plate region. In addition, the license plate type is identified
by the difference in the gray-level value between the background region and the character region. The proposed method is effective
in solving the current challenges in extracting the license plate region from the damaged frames of license plates issued
for domestic use, including new types of license plates. According to the experimental results, the proposed method yields
a high extraction rate of 99.5% for vehicle license plates. 相似文献
12.
13.
M. -K. Kim 《International Journal of Automotive Technology》2010,11(5):751-758
License plate location is a challenging task that is necessary for automatic vehicle identification. This paper presents a
new method for locating a license plate when its size and aspect ratio are highly variable. The proposed method begins with
an assumption that a license plate exists in a region where dense edges are located. We define an edge region as an area containing
rich edges. The edge regions are created by dilating vertical edges, and they are classified into one of four types: left
fragment type, right fragment type, whole type, and undefined type. The candidates for a license plate region are constructed
by merging edge regions. Knowing what type of edge region is being examined is useful in the merging process. Finally, we
verify whether each candidate contains a license plate or not by using the character arrangement information. The arrangement
pattern is determined by the size of connected components and by the vertical overlap or horizontal distance between two neighboring
components. Experimental results show that the proposed method gives robust results regardless of any variation in the size
and aspect ratio of license plates. 相似文献
14.
车牌定位是车牌自动识别系统的关键。提出了一种以纹理特征为主要手段,以彩色特征为辅助手段的多特征定位算法,对不同背景和光照条件下的各类车牌图像进行了实验,结果表明该算法是一种可行的方法。 相似文献
15.
16.
车牌定位及车辆识别是智能交通管理的主要研究问题.车牌定位识别,通过对图像进行预处理并结合形态学能粗略获取候选车牌位置,对符合特征的候选车牌进行筛选,精确获取车牌位置,最后采用神经网络完成字符识别过程.车辆识别采用迁移学习,采用AlexNet卷积神经网络构造出深度特征向量.形态学能够应对灰度底质量差的情形,为字符识别提供保障.车辆识别时对比直接分类图片特征,迁移学习构造的深度特征分类精度为85.13%,提高了38%,验证了迁移学习的有效性,通过KNN算法表明深度特征能够表征图片属性.针对新数据集重新提取特征、训练样本将消耗大量时间,对比迁移学习和AlexNet框架发现分类精度持平,表明了迁移学习的鲁棒性. 相似文献