| 摘 要: 自动车牌识别技术在智能交通系统中发挥着重要作用，针对智能手持设备拍照分辨率低、角度不定等问题，设计和优化了车牌识别中的步骤，包括：(1)基于形态学的车牌定位；(2)基于边缘检测的车牌倾斜校正；(3)基于人工神经网络的车牌字符分割与识别等步骤。对手机拍摄的车牌照片进行实验，所提出的算法利用神经网络匹配识别精度达到95.2%，平均运行时间为2.015 s，无论识别精度还是时间都能够达到应用需求。 |
| 关键词: 智能手持设备；车牌定位；边缘检测；人工神经网络；车牌识别 |
|
中图分类号: TP311.5
文献标识码: A
|
| 基金项目: 枣庄学院博士科研启动基金(1020714). |
|
| Design of License Plate Location and Recognition System based on Handheld Device Image |
|
SUN Peng1, LI Sai1, KOU Peng2, ZHU Jiajun1, HAN Xirui1, ZHANG Tianyi3
|
( 1.School of mechanical and electrical engineering, Zaozhuang University, Zaozhuang 277160, China;
2.Liupanshui Natural Resources Bureau, Liupanshui 553099, China;
3.School of tourism, resources and environment, Zaozhuang University, Zaozhuang 277160, China )
zzxyjdy032@163.com; 814100090@qq.com; 1298950858@qq.com; zzxyjdy045@163.com; zzxyjdy078@163.com; zzxylz025@163.com
|
| Abstract: Automatic license plate recognition technology plays an important role in intelligent transportation systems. Aiming at the problems of low resolution and variable angles of smart handheld devices, this paper proposes to design and optimize the steps in license plate recognition. The optimized steps include (1) Morphology-based license plates location; (2) License plate tilt correction based on edge detection; (3) License plate character segmentation and recognition based on artificial neural network. Experiments on license plate photos taken by mobile phones show that the proposed algorithm achieves a recognition accuracy of 95.2%, and an average running time of 2.015 s, which can meet the application requirements of both the recognition accuracy and time. |
| Keywords: smart handheld device; license plate positioning; edge detection; artificial neural network; license plate location |
