基于色彩空间的新能源汽车识别方法

New energy vehicle recognition method based on color space

为适应汽车使用结构的变化,越来越多的停车场设有新能源汽车停车位,并为其提供充电服务。但由于车位紧张,该停车位常被燃油车占用,导致新能源汽车无法充电。针对这一问题,提出了一种基于色彩空间的新能源汽车识别方法。该方法不仅能正确定位和识别车牌号,还能对车辆类型进行分类,是解决新能源汽车停车难问题的有效方法。首先,采用车牌几何特征与颜色特征相结合的方法对车牌进行定位。然后,通过色调饱和度明度(HSV)色彩空间识别出车辆类型。最后,采用基于残差网络(ResNet)模型的字符识别方法得到车牌号。实验表明,该方法车辆类型识别正确率为96.6%,车牌号识别正确率为97.53%。

In order to face the changes in the vehicle structure,more and more car parks set up parking spaces for new energy vehicles and provide charging services.However,the fuel vehicles usually park in the spaces for new energy vehicles,which causes the inability of new energy vehicles to charge.To solve this problem,a new energy vehicle recognition method is proposed based on color space.This method can not only correctly identify the license plate number,but also identify and classify the vehicle type.Therefore,this method is an effective method to solve the difficult problem of new energy vehicle parking.First,the geometric and color feature are combined to locate the license plate,and then the vehicle model is identified through the HSV color space.finally,the license plate number is obtained by the ResNet model.The experiments show that the recognition rates of the vehicle model and license plate number by the proposed method are 96.6%and 97.53%,respectively.