[1]刘 彬,严京旗,施鹏飞.高斯差分的AdaBoost车牌定位方法[J].智能系统学报,2010,5(6):471-475.
LIU Bin,YAN Jing-qi,SHI Peng-fei.A license plate location method based on the DoG+AdaBoost algorithm[J].CAAI Transactions on Intelligent Systems,2010,5(6):471-475.
点击复制
LIU Bin,YAN Jing-qi,SHI Peng-fei.A license plate location method based on the DoG+AdaBoost algorithm[J].CAAI Transactions on Intelligent Systems,2010,5(6):471-475.
高斯差分的AdaBoost车牌定位方法
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
5
期数:
2010年第6期
页码:
471-475
栏目:
学术论文—机器感知与模式识别
出版日期:
2010-12-25
- Title:
- A license plate location method based on the DoG+AdaBoost algorithm
- 文章编号:
- 1673-4785(2010)06-0471-05
- 关键词:
- 复杂光照条件; 车牌定位; AdaBoost算法; 高斯差分
- Keywords:
- complex illumination conditions; license plate location; AdaBoost algorithm; differential of Gaussian (DoG)
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 针对实际拍摄场景中的复杂光照条件以及不同车牌颜色对车牌定位造成的影响，提出了一种基于高斯差分图像的AdaBoost车牌检测算法.该算法首先对原始灰度图像进行高斯差分，得到其对应的高斯差分图像，然后利用基于DoG图像的DoG+AdaBoost分类器与基于灰度图像的Gray+AdaBoost分类器构成二级车牌检测器进行车牌检测，最后根据车牌中的车牌号码信息对车牌检测结果进行验证，得到最终的车牌定位结果.该算法利用高斯差分方法，很好地抑制了复杂光照和不同车牌颜色对车牌检测造成的影响，具有较快的定位速度和很高的检出率.实验表明，该算法能获得很好的车牌定位效果，具有较高的实用价值.
- Abstract:
- While focusing on the effect of complex illumination conditions in an actual shooting scene and different colors of a license plate in various locations, a license plate detection method was presented based on the DoG+AdaBoost algorithm. First, by differential Gaussian processing, the DoG image was obtained from the grayscale image, and then a 2stage detector formed by the DoG+AdaBoost classifier and Gray+AdaBoost classifier was used to detect the license plate. Finally, the registration number information was used to verify the license plate and to produce the final result. The algorithm greatly reduced the effect of complex illumination conditions and different colors of the license plate in various locations with a fast processing time and high detection rate. The algorithm performed well in the experiment and possesses high practical value.
- 参考文献/References:
-
［1］HONGLIANG B， CHANGPING L. A hybrid licenseplate extraction method based on edge statistics and morphology［C］//ICPR.［S.l.］， 2004： 831834.
［2］LLORENS D, MARZAL A, PALAZON V, et al. Car license plates extraction and recognition based on connected components analysis and HMM decoding［C］// New York: SpringerVerlag, 2005, 3522: 571578.
［3］BROUMANDNIA A, FATHY M. Application of pattern recognition for Farsi license plate recognition［C］//GVIP. Cairo, Egypt, 2005： 623627.
［4］郭捷, 施鹏飞. 基于颜色和纹理分析的车牌定位方法［J］.中国图象图形学报, 2002, 7 (5) : 472476.
GUO Jie, SHI Penfei. Color and texture analysis based vehicle license plate location［J］. Journal of Image and Graphics, 2002， 7(5): 472476.
［5］DLAGNEKOV L. License plate detection using AdaBoost［D］. San Diego, USA: University of California, 2004.
［6］盛曦, 吴炜. 一种基于AdaBoost的车牌定位算法［J］.四川大学学报, 2007, 44(3): 535539.
SHENG Xi, WU Wei. A license plate location method based on Adaboost algorithm［J］. Journal of Sichuan University, 2007, 44 (3): 535539.
［7］LOWE D G. Distinctive image features from scaleinvariant keypoints［J］. International Journal of Computer Vision, 2004, 60(2): 91110.
［8］FREUND Y, SCHAPIRE R E. A decisiontheoretic generalization of online learning and an application to boosting［C］//Computational Learning Theory: Eurocolt〖KG-*1/3〗 ’95. ［S.l.］， 1995: 2337.
［9］VIOLA P, JONES M. Rapid object detection using a boosted cascade of simple features［C］// Conference on Computer Vision and Pattern Recognition. Kauai, USA: IEEE Press, 2001: 511518.
- 相似文献/References:
-
[1]黄 腾,施鹏飞.非均匀光照下的车牌定位方法[J].智能系统学报,2009,4(5):421.[doi:10.3969/j.issn.1673-4785.2009.05.006]
HUANG Teng,SHI Peng-fei.Locating vehicle license plates in a non-uniformly illuminated environment[J].CAAI Transactions on Intelligent Systems,2009,4():421.[doi:10.3969/j.issn.1673-4785.2009.05.006]
[2]吴加明,吴一全.基于Tent映射CPSO和车牌纹理特征的车牌定位[J].智能系统学报,2011,6(4):333.
WU Jiaming,WU Yiquan.License plate location based on texture features and tent chaotic particle swarm optimization[J].CAAI Transactions on Intelligent Systems,2011,6():333.
备注/Memo
收稿日期：2010-04-17.
基金项目：国家自然科学基金资助项目(60873137).
通信作者：严京旗. E-mail:jqyan@sjtu.edu.cn.
作者简介：
刘 彬，男，1984年生，硕士研究生，主要研究方向为图像处理、机器学习与生物特征识别.
?严京旗，男，1975年生，副教授，硕士生导师，主要研究方向为图像图形综合技术、可视计算、三维生物特征识别等.曾多次赴香港理工大学计算机系开展合作研究并建立了密切的合作关系.参与并主持了多项国家自然科学基金项目.
?施鹏飞，男，1940年生，教授，博士生导师.主要研究方向为图像处理、模式识别及智能技术与系统.国家自然科学基金信息学科、国家教育部、科技部评审与咨询专家.国际IEEE高级会员，中国人工智能学会常务理事，中国科学院自动化所模式识别国家重点实验室、北京大学机器感知与智能教育部国家重点实验室学术委员.获国家、省部级科技奖6项，发表学术论文100余篇.
更新日期/Last Update:
2011-03-03