基于颜色特征和边缘检测的车牌识别算法

License Plate Recognition System Based on Color Features and Edge Detection

完整的车牌识别系统由车牌定位、字符分割和字符识别三部分组成,针对车牌定位和字符分割两项核心技术,对传统算法进行了改进,提出了一种新的车牌定位和字符分割的算法,解决了传统算法易受噪声干扰且识别率不高的问题。该算法利用颜色特征和边缘检测技术实现对车牌的精准定位,并利用倾斜校正和垂直扫描技术实现了对字符的分割,最后结合神经网络的字符识别技术形成一个完整的车牌识别系统。利用复杂的环境条件下采集的车辆图像对该系统进行了大量实验,实验结果显示识别率达到97%,证明了算法鲁棒性好,识别率高。

A complete license plate recognition system consists of license plate location,character segmentation and character recognition.Aiming at the two core technologies of license plate location and character segmentation,this paper improves the traditional algorithm and proposes a new algorithm for license plate location and character segmentation,which solves the problem that the traditional algorithm is easy to be disturbed by noise and has low recognition rate.The algorithm uses color features and edge detection technology to achieve accurate location of license plate,and uses tilt correction and vertical scanning technology to achieve the segmentation of characters,finally,combined with the character recognition technology of neural network,a complete license plate recognition system is formed.In this paper,a large number of experiments are carried out on the system with vehicle images collected under complex environmental conditions.The experimental results show that the recognition rate reaches 97%,which proves that the algorithm has good robustness and high recognition rate.