陕西理工学院学报(自然科学版)
陝西理工學院學報(自然科學版)
협서리공학원학보(자연과학판)
JOURNAL OF SHAANXI UNIVERSITY OF TECHNOLOGY (NATURAL SCIENCE EDITION)
2014年
4期
27-31
,共5页
车牌识别%低频融合%支持向量机
車牌識彆%低頻融閤%支持嚮量機
차패식별%저빈융합%지지향량궤
license plate recognition%low frequency fusion%support vector machines
针对目前市场上使用的监控设备在雾霾天气下辨识精度和速度不足等问题,采用小波低频融合原理对采集到的图像预处理,并基于改进支持向量机算法对预处理后的车牌图像提取特征值。实验结果表明,改进后的车牌识别方法与传统方法相比,提高了辨别速度,减小了误识率,为雾天视频监控系统的研发提供了新的思路。
針對目前市場上使用的鑑控設備在霧霾天氣下辨識精度和速度不足等問題,採用小波低頻融閤原理對採集到的圖像預處理,併基于改進支持嚮量機算法對預處理後的車牌圖像提取特徵值。實驗結果錶明,改進後的車牌識彆方法與傳統方法相比,提高瞭辨彆速度,減小瞭誤識率,為霧天視頻鑑控繫統的研髮提供瞭新的思路。
침대목전시장상사용적감공설비재무매천기하변식정도화속도불족등문제,채용소파저빈융합원리대채집도적도상예처리,병기우개진지지향량궤산법대예처리후적차패도상제취특정치。실험결과표명,개진후적차패식별방법여전통방법상비,제고료변별속도,감소료오식솔,위무천시빈감공계통적연발제공료신적사로。
In view of inaccruacy and inadequacy in speed of recoginition of the current monitoring e-quipment in smoggy weather,the study uses low-frequency wavelet fusion theory to pretreat the collected. Then,post-treatment is administered to extract distinguished features of the license plate based on improved support vector machines. The experimental results show that:compared with the traditional methods for vehicle license plate recognition,the proposed method improves the speed discrimination,reduces the misrecognition and provides a new way for video monitor system developers in the haze weather.