边缘检测算法在车牌定位系统中的应用研究
|
摘要
汽车牌照识别系统是近几年发展起来的基于图像处理和字符识别技术的智能化交通管理系统。它可以广泛地应用于交通监测和管理部门中,能够大大提高工作效率。车牌识别技术主要包括车牌定位和车牌识别两部分。而车牌定位是该系统的关键之一。由于牌照拍摄场景的复杂性以及车牌位置和图像质量的不可预知性,车牌定位系统一直都未做到令人满意,所以车牌定位一直是这个领域研究的热点。具有很高的经济价值和实用价值。
本文首先对常见的图像处理方法进行研究和实验。在车牌定位系统中,我们用两个模块分别对图像预处理、图像的边缘检测和分割的若干方法进行编程实验。并根据系统的需要进行了必要的改进,提出了一维垂直边缘检测算子,该算子能够很好的突出车牌区域特征。
其次,提出了基于垂直边缘检测算子的车牌定位方法,根据汽车牌照区域的垂直边缘统计特性,从图像中确定可能存在的牌照候选区,在利用车牌几何形状上的特点对这些候选区进行筛选,得到车牌位置。该算法对采集的图像数据库的实验效果较好。
然后,利用图像分割法和形态学中的腐蚀法,在系统中设计了车牌归一化和车牌细化模块。实现了从车牌定位到车牌识别的连接。
最后,在上述工作成果的基础上,利用Visual C++编译环境,设计了可独立运行的车牌定位系统软件,该软件具有可视化、模块化、使用简便的特点,并且运行效果良好。
Based on the technology of recognizing images and characters, the vehicle license plate recognition system is a kind of intelligent transportation management system. Such systems, which are applied in parking areas, highways, bridges and tunnels, can help a human operator and improve the overall quality of a service. Automatic detection and recognition of vehicle license plates mainly includes the subsystem of license plate location and character recognition. The location of license plates is the key of the system. Because of the complex of image background, the uncertainty of plate position and image quality, the location of plates is not satisfied. Therefore, the study on the algorithms of license plate location is always the hotspot problem. The system has very big economy value and practicality.
Above all, familiar image processing methods were processed analyzing and experiment. In the system of license plate location, we used two modules block answer the image preprocessing and image of edge detection and intersected, programme and experiment were done for several methods. Based on the need of system, we carried through essential improvement, proposed a new algorithm for perpendicular edge detection, it can beautiful extrude license plate provincial characteristics.
Secondly, in this paper a new arithmetic of perpendicular edge detection work on license plate location, according to statistical features of perpendicular edge which belong to license plate provincial characteristics, from the image, it can confirm the likelihood license plate inquire electoral district. Utilization the speciality of geometrical shape, we can filter the license plate inquire electoral district, then located the license plate location. The arithmetic had a very good result for experiment in the image database.
Afterward, availing image cutting method and morphologic’s etching method, in the system, we designed two model blocks normalization and thinning. Finishing the connecting from location to discriminating.
At last, according to the above research, license plate location system software is developed in the environment of Visual C++. The software is characterized by modularization, visualization and simple use. Result of this software is good.
本文首先对常见的图像处理方法进行研究和实验。在车牌定位系统中,我们用两个模块分别对图像预处理、图像的边缘检测和分割的若干方法进行编程实验。并根据系统的需要进行了必要的改进,提出了一维垂直边缘检测算子,该算子能够很好的突出车牌区域特征。
其次,提出了基于垂直边缘检测算子的车牌定位方法,根据汽车牌照区域的垂直边缘统计特性,从图像中确定可能存在的牌照候选区,在利用车牌几何形状上的特点对这些候选区进行筛选,得到车牌位置。该算法对采集的图像数据库的实验效果较好。
然后,利用图像分割法和形态学中的腐蚀法,在系统中设计了车牌归一化和车牌细化模块。实现了从车牌定位到车牌识别的连接。
最后,在上述工作成果的基础上,利用Visual C++编译环境,设计了可独立运行的车牌定位系统软件,该软件具有可视化、模块化、使用简便的特点,并且运行效果良好。
Based on the technology of recognizing images and characters, the vehicle license plate recognition system is a kind of intelligent transportation management system. Such systems, which are applied in parking areas, highways, bridges and tunnels, can help a human operator and improve the overall quality of a service. Automatic detection and recognition of vehicle license plates mainly includes the subsystem of license plate location and character recognition. The location of license plates is the key of the system. Because of the complex of image background, the uncertainty of plate position and image quality, the location of plates is not satisfied. Therefore, the study on the algorithms of license plate location is always the hotspot problem. The system has very big economy value and practicality.
Above all, familiar image processing methods were processed analyzing and experiment. In the system of license plate location, we used two modules block answer the image preprocessing and image of edge detection and intersected, programme and experiment were done for several methods. Based on the need of system, we carried through essential improvement, proposed a new algorithm for perpendicular edge detection, it can beautiful extrude license plate provincial characteristics.
Secondly, in this paper a new arithmetic of perpendicular edge detection work on license plate location, according to statistical features of perpendicular edge which belong to license plate provincial characteristics, from the image, it can confirm the likelihood license plate inquire electoral district. Utilization the speciality of geometrical shape, we can filter the license plate inquire electoral district, then located the license plate location. The arithmetic had a very good result for experiment in the image database.
Afterward, availing image cutting method and morphologic’s etching method, in the system, we designed two model blocks normalization and thinning. Finishing the connecting from location to discriminating.
At last, according to the above research, license plate location system software is developed in the environment of Visual C++. The software is characterized by modularization, visualization and simple use. Result of this software is good.
引文
1 郑南宁等.行驶车辆牌照自动识别系统. 西安交通大学报. 1991(1): 43-54
2 郁 梅等.基于视觉的车辆牌照检测.计算机应用研究. 1999(5): 65-67
3 沈定刚等.汽车牌照自动定位算法.上海交通大学光纤技术研究所. 1995, 05: 35-38
4 袁宝民等.汽车牌照定位研究综述.大连大学学报.2002, 4: 6-12
5 Lv Xuefeng. Wang Min. Wan Guohong. Peng Gang. Study on License Plate Segmentation Huazhong University of Science and Technology. Wuhan 430074
6 Paolo Ferragina, Mario Notturno Graieri, Flavio Stabile.Optical Recognition of Motor Vehicle License Plate. Paolo Comelli IEEE Transactions on Vehicle Technology, 1995, 44(4): 790-799
7 徐全生等. 汽车牌照图像的预处理研究. 沈阳工业大学学报. 2002, 4: 121-124
8 Lee. Eun Ryung. kim, Pyeoung Kee Kim, Hang Joon, Automatic recognition of a vehicle license plate using color image processing, IEEE International Conference on Image Processing 1 Nov 13-16 2001:301-305
9 Park, S.H Kim, K.I.Jung, K. Kin, H.J., Location vehicle license plates using neural Networks, Electronics Letters v 35 n 17 2002:1475-1477
10 金玲玲等. 汽车牌照的提取方法. 南理工大学学报(自然科学版) .2002, 7: 95-98
11 Luis Salgado, Jose M.Memendez, Enrique Rendon, Narciso Garcia. Automatic Car Plate Detection and Recognition through Intelligent Vision Engineering. IEEE Annual International Carnahan Conference on Security Technology. Proceedings, 1999:71-76
12 董慧颖等. 汽车牌照自动识别系统总体设计及软件结构. 沈阳工业学院学报. 2003, 9: 27-29
13 景云华等. 数字图像处理技术在车辆自动识别系统中的应用. 计算机工程. 2003, 9:1-3
14 陆艺等. 图像处理技术在机动车检测行业中的应用. 中国测试技术. 2003, 11: 36-37
15 杨枝灵. 王开等. Visual C++数字图像获取处理及实践应用. 人民邮政出版社 2003, 4:206-218
16 Mallat S, Zhong S. Characterization of signal from multiscale edge [J]. IEEE Trans PAMI. 1992, (7):710-732.
17 Jeong H. Kim C. I. Adaptive determination of filter scales for edge detection. IEEE Transactions on Pattern Analysis Machine Intelligence.1992, 14(5):579-585.
18 Bergholm F. Edge. focusing. IEEE Transactions on Pattern Analysis Machine Intelligence. 1987, 9(6): 726-741.
19 韩永强. 李世祥. 汽车牌照子图像的定位算法. 微型电脑应用. 1999(3): 14-16
20 郁 梅等. 基于视觉的车辆牌照检测. 计算机应用研究. 1999 (5) : 65-67
21 李毅. 改进型公路车辆监控系统及其应用研究. 北京大学硕士学位论文. 1999: 1-2
22 林惠保. 车号识别算法的研究与系统的实现. 北京大学硕士学位论文. 2000: 30-35
23 Yuntao Cui, Qian Huang. Extracting characters of license plates from video sequences, Machine Vision and Applications, vol.10(5-6)2004: 308-320
24 Kenneth R. Castleman. Digital image processing. Prentice Hall International. Inc, 2001: 201-220
25 季梁. 数字图像处理. 北京:清华大学出版社. 2002: 120-130
26 Kamat. Varsha and Ganesan. Subramanian Efficient implementation of the Hough Transform for detecting vehicle license plates using DSP’s. Real-Time Technology and Applications-Proceedings May 15-17 2003 : 58-59
27 S.Messelodi, C.M. Modena. Automatic identification and skew estimatition of text lines inreal scene images. Pattern Recognition 32(1999): 791-810
28 Viola P. Jones M. Repid object detection using a boosted cascade of simple features. Computer Vision and Pattern Recognition. 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on, Volume:1, 8-14 Dec.2001: 511-518
29 乐宁, 翁世修.基于单字符区域的倾斜字符校正技术研究[J].图像识别与自 动化. 2002, (2):1-7
30 胡爱明, 周孝宽. 利用形态特征的汽车车牌图像分割方法[J]. 计算机辅助设计与图形学报. 2003, 15(6): 716-719
31 李文举, 梁德群, 崔连延, 毕胜.一种新的面向字符分割的车牌图像预处理 方法. 计算机应用研究.2004(7): 258-260
32 马颂德, 张正友著. 计算机视觉-计算理论与算法基础. 北京:科学出版社. 1998: 11-12
33 叶子青著. Visual C++系统开发实例精粹. 人民邮电出版社. 2005: 97-245
34 古槿. 新手学 Visual C++30 例. 人民邮电出版社. 2004: 30-35
35 T. N. Tan. Efficient Image Gradient Based Vehicle Localization. IEEE Transportations on Image Processing. Auguest 2000,9: 1343-1356
36 庭芝, 方子文著. 数字图像处理及模式识别. 北京理工大学出版社. 1997, 8: 180-190
37 牛欣. 汽车牌照识别技术研究. 测控技术. 1999,18(12): 14-17
38 廖金周, 宜国荣. 车辆牌照的自动分割 .微型电脑应用. 1999(7):32-34
39 J. KITTLER, J. ILLINGWORTH. Minimum Error Thresholding. Pattern Recognition, 1986, 19(1): 41-47
40 stu N. A threshold selection method from gray-level histogram. IEEE Trans, 1979: SMC-9: 62-66
41 引. 潘云鹤. 彩色汽车图像牌照定位新方法. 中国图形图像学报.2001, 6: 35-36
42 郭捷. 施鹏飞. 基于颜色和纹理分析的车牌定位方法.中国图像图形学报. Vol, 7(A) 2002, 5: 472-476
43 明亮. 范勇. 一种快速精确的车辆牌照字符切分算法. 四川大学学报(自然科学版). 2002, 39(6): 50-53
44 刚. 丁晓青. 多知识综合判决字符切分算法. 计算机工程与应用. 2002, Vol. 17: 59-62
45 陈黎. 黄心汉. 基于聚类分析的车牌字符切分方法. 计算机工程与应用. 2002, 6: 221-256
46 何为,韩力群. 汽车牌照自动识别技术的现状与发展[J]. 北京轻工业学院学报. 2001,19(1): 36-39
47 心明, 兰赛, 徐燕. 图像处理中几种边缘检测算法的比较[J].现代电力. 2000,17(3): 65-69
48 陈展, 王敏, 黄心汉.一种基于边缘特征的汽车牌照定位算法[J].华中科技大学学报(自然科学版). 2004 (10) : 98-99
49 曾致远, 付祥胜. 基于数学形态学和边缘特征的车牌定位算法[J]. 电视技术. 2005 (7) : 94-96
50 丁秋芳. 车牌自动定位切分技术. 江苏交通. 2003(1): 23-24
2 郁 梅等.基于视觉的车辆牌照检测.计算机应用研究. 1999(5): 65-67
3 沈定刚等.汽车牌照自动定位算法.上海交通大学光纤技术研究所. 1995, 05: 35-38
4 袁宝民等.汽车牌照定位研究综述.大连大学学报.2002, 4: 6-12
5 Lv Xuefeng. Wang Min. Wan Guohong. Peng Gang. Study on License Plate Segmentation Huazhong University of Science and Technology. Wuhan 430074
6 Paolo Ferragina, Mario Notturno Graieri, Flavio Stabile.Optical Recognition of Motor Vehicle License Plate. Paolo Comelli IEEE Transactions on Vehicle Technology, 1995, 44(4): 790-799
7 徐全生等. 汽车牌照图像的预处理研究. 沈阳工业大学学报. 2002, 4: 121-124
8 Lee. Eun Ryung. kim, Pyeoung Kee Kim, Hang Joon, Automatic recognition of a vehicle license plate using color image processing, IEEE International Conference on Image Processing 1 Nov 13-16 2001:301-305
9 Park, S.H Kim, K.I.Jung, K. Kin, H.J., Location vehicle license plates using neural Networks, Electronics Letters v 35 n 17 2002:1475-1477
10 金玲玲等. 汽车牌照的提取方法. 南理工大学学报(自然科学版) .2002, 7: 95-98
11 Luis Salgado, Jose M.Memendez, Enrique Rendon, Narciso Garcia. Automatic Car Plate Detection and Recognition through Intelligent Vision Engineering. IEEE Annual International Carnahan Conference on Security Technology. Proceedings, 1999:71-76
12 董慧颖等. 汽车牌照自动识别系统总体设计及软件结构. 沈阳工业学院学报. 2003, 9: 27-29
13 景云华等. 数字图像处理技术在车辆自动识别系统中的应用. 计算机工程. 2003, 9:1-3
14 陆艺等. 图像处理技术在机动车检测行业中的应用. 中国测试技术. 2003, 11: 36-37
15 杨枝灵. 王开等. Visual C++数字图像获取处理及实践应用. 人民邮政出版社 2003, 4:206-218
16 Mallat S, Zhong S. Characterization of signal from multiscale edge [J]. IEEE Trans PAMI. 1992, (7):710-732.
17 Jeong H. Kim C. I. Adaptive determination of filter scales for edge detection. IEEE Transactions on Pattern Analysis Machine Intelligence.1992, 14(5):579-585.
18 Bergholm F. Edge. focusing. IEEE Transactions on Pattern Analysis Machine Intelligence. 1987, 9(6): 726-741.
19 韩永强. 李世祥. 汽车牌照子图像的定位算法. 微型电脑应用. 1999(3): 14-16
20 郁 梅等. 基于视觉的车辆牌照检测. 计算机应用研究. 1999 (5) : 65-67
21 李毅. 改进型公路车辆监控系统及其应用研究. 北京大学硕士学位论文. 1999: 1-2
22 林惠保. 车号识别算法的研究与系统的实现. 北京大学硕士学位论文. 2000: 30-35
23 Yuntao Cui, Qian Huang. Extracting characters of license plates from video sequences, Machine Vision and Applications, vol.10(5-6)2004: 308-320
24 Kenneth R. Castleman. Digital image processing. Prentice Hall International. Inc, 2001: 201-220
25 季梁. 数字图像处理. 北京:清华大学出版社. 2002: 120-130
26 Kamat. Varsha and Ganesan. Subramanian Efficient implementation of the Hough Transform for detecting vehicle license plates using DSP’s. Real-Time Technology and Applications-Proceedings May 15-17 2003 : 58-59
27 S.Messelodi, C.M. Modena. Automatic identification and skew estimatition of text lines inreal scene images. Pattern Recognition 32(1999): 791-810
28 Viola P. Jones M. Repid object detection using a boosted cascade of simple features. Computer Vision and Pattern Recognition. 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on, Volume:1, 8-14 Dec.2001: 511-518
29 乐宁, 翁世修.基于单字符区域的倾斜字符校正技术研究[J].图像识别与自 动化. 2002, (2):1-7
30 胡爱明, 周孝宽. 利用形态特征的汽车车牌图像分割方法[J]. 计算机辅助设计与图形学报. 2003, 15(6): 716-719
31 李文举, 梁德群, 崔连延, 毕胜.一种新的面向字符分割的车牌图像预处理 方法. 计算机应用研究.2004(7): 258-260
32 马颂德, 张正友著. 计算机视觉-计算理论与算法基础. 北京:科学出版社. 1998: 11-12
33 叶子青著. Visual C++系统开发实例精粹. 人民邮电出版社. 2005: 97-245
34 古槿. 新手学 Visual C++30 例. 人民邮电出版社. 2004: 30-35
35 T. N. Tan. Efficient Image Gradient Based Vehicle Localization. IEEE Transportations on Image Processing. Auguest 2000,9: 1343-1356
36 庭芝, 方子文著. 数字图像处理及模式识别. 北京理工大学出版社. 1997, 8: 180-190
37 牛欣. 汽车牌照识别技术研究. 测控技术. 1999,18(12): 14-17
38 廖金周, 宜国荣. 车辆牌照的自动分割 .微型电脑应用. 1999(7):32-34
39 J. KITTLER, J. ILLINGWORTH. Minimum Error Thresholding. Pattern Recognition, 1986, 19(1): 41-47
40 stu N. A threshold selection method from gray-level histogram. IEEE Trans, 1979: SMC-9: 62-66
41 引. 潘云鹤. 彩色汽车图像牌照定位新方法. 中国图形图像学报.2001, 6: 35-36
42 郭捷. 施鹏飞. 基于颜色和纹理分析的车牌定位方法.中国图像图形学报. Vol, 7(A) 2002, 5: 472-476
43 明亮. 范勇. 一种快速精确的车辆牌照字符切分算法. 四川大学学报(自然科学版). 2002, 39(6): 50-53
44 刚. 丁晓青. 多知识综合判决字符切分算法. 计算机工程与应用. 2002, Vol. 17: 59-62
45 陈黎. 黄心汉. 基于聚类分析的车牌字符切分方法. 计算机工程与应用. 2002, 6: 221-256
46 何为,韩力群. 汽车牌照自动识别技术的现状与发展[J]. 北京轻工业学院学报. 2001,19(1): 36-39
47 心明, 兰赛, 徐燕. 图像处理中几种边缘检测算法的比较[J].现代电力. 2000,17(3): 65-69
48 陈展, 王敏, 黄心汉.一种基于边缘特征的汽车牌照定位算法[J].华中科技大学学报(自然科学版). 2004 (10) : 98-99
49 曾致远, 付祥胜. 基于数学形态学和边缘特征的车牌定位算法[J]. 电视技术. 2005 (7) : 94-96
50 丁秋芳. 车牌自动定位切分技术. 江苏交通. 2003(1): 23-24
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |