联合视觉与规则评价的车牌放大号检测识别方法

Detection and Recognition Method of Enlarged License Plate Based on Combined Vision and Rule Evaluation

大型车辆的身份识别在交通系统中十分重要,但由于牌照位置、拍摄角度以及污渍的影响,目前的车牌检测方法并不能很好的得到其牌照信息.而通过检测车厢后部喷涂的车牌放大号对其车牌号码进行获取是解决这一问题的有效途径,由于车牌放大号的字符间距比较大,当前的文本检测方法容易出现误检,因此本文提出一种联合视觉与规则评价的车牌放大号检测识别方法.利用注意力模块对特征进行增强,构建整体检测、掩膜检测与字符检测识别的多分支网络对车牌放大号整体与字符分别进行获取,同时针对车牌放大号的字符特点与命名规则设计评价模块.将评价模块的评价分数与整体检测的置信度分数进行联合评价.经实验表明,该方法对车牌放大号的检测精准率达到了96.34%,验证了该方法的可行性.

The identification of large vehicles is very important in the traffic system,but due to the influence of license plate location,shooting angle and stains,the current license plate detection methods can not get its license plate information very well.However,obtaining the license plate number by detecting the enlarged license plate painted at the rear of the truck is an effective way to solve this problem.Because the character spacing of the enlarged license plate is relatively large,the current text detection methods are prone to false detection.Therefore,this paper proposes a detection and recognition method of enlarged license plate based on combined vision and rule evaluation.The attention module is used to enhance the features,and the multi-branch network of the whole detection,the mask detection and the character detection recognition is constructed to obtain the whole and characters of the enlarged license plate respectively.At the same time,an evaluation module is designed according to the character characteristics and naming rules of the enlarged license plate.The evaluation score of the evaluation module and the confidence score of the whole detection module are evaluated jointly.The experimental results show that the detection accuracy of this method for enlarged license plate reaches 96.34%,which verifies the feasibility of this method.