CAJ | 학술논문

针对车辆图像分割比较困难、车牌位置定位非常不准等问题，提出了一种新的车牌快速定位方法。在对图像作灰度化、灰度拉伸、边缘检测和二值化等预处理的基础上，根据车牌区域纹理丰富的特点，对得到的二值图像作行、列扫描，并利用车牌位置信息、尺寸信息来缩小搜索范围，提高了定位精度和实时性。对不同光照、不同场景、不同大小的汽车图像的仿真实验结果表明，该方法具有较好的定位精度和鲁棒性，实时性强，适用于不同的应用场合。
침대차량도상분할비교곤난、차패위치정위비상불준등문제，제출료일충신적차패쾌속정위방법。재대도상작회도화、회도랍신、변연검측화이치화등예처리적기출상，근거차패구역문리봉부적특점，대득도적이치도상작행、렬소묘，병이용차패위치신식、척촌신식래축소수색범위，제고료정위정도화실시성。대불동광조、불동장경、불동대소적기차도상적방진실험결과표명，해방법구유교호적정위정도화로봉성，실시성강，괄용우불동적응용장합。
In terms of difficulty in vehicle image segmentation and inaccuracy in vehicle positioning, the paper puts forward a new method to quickly locate the binaryzation, the paper carries out row and texture of license plate area. Then it narro accuracy and timeliness of positioning. Mod and different sizes show that the method is vehicles. Based on such pre--processes as image graying, gray expanding and column scanning of the resultant binary images according to the features of rich ws down the hunting zone of vehicle license position and size to enhance the el experimental imaging results of vehicles in different lighting, different scenes highly, accurate and robust, timely and feasible to different occasions.