基于遗传算法的车牌定位研究

Research on Lience Plate Location Based on Genetic Algorithm

车牌识别过程中,车牌定位起着重要作用.采用基于颜色的方法对车牌进行定位时,结果显示图像下边缘丢失,不能正确定位车牌,原因是未找到最优阈值.对此,尝试提出一种自适应变异算子的改进遗传算法,该自适应变异概率改进遗传算法可以随着进化代数自适应改变,在前期具有更好的全局搜索能力,在后期也拥有更高的收敛精度.经过两个典型测试函数性能测试,改进遗传算法的有效性高于标准遗传算法.将改进遗传算法应用到车牌定位中,能够实现车牌的准确定位.

In the process of license plate recognition,license plate location plays an important role.When using the color based method to locate the license plate,the result shows that the lower edge of the image is lost and the license plate cannot be located correctly,because the optimal threshold is not found.In this paper,an improved genetic algorithm of adaptive mutation operator is proposed.The improved genetic algorithm of adaptive mutation probability can change adaptively with the evolution algebra,which has better global search ability in the early stage and higher convergence accuracy in the later stage.After the performance test of two typical test functions,the effectiveness of the improved genetic algorithm is higher than that of the standard genetic algorithm.Applying the improved genetic algorithm to the license plate location can realize the accurate location of the license plate.