摘要
为了能够实时采集铁轨图像以实现陆轨两用车在站台与铁轨之间顺利切换,介绍了一种基于多尺度小波变换和自适应阈值的图像处理方法,选用二次B样条小波对图像传感器采集到的铁轨图像进行多尺度边缘检测,结合了非极大值抑制方法和自适应阈值方法提取最终边缘图像。对比几种经典的边缘检测算子,小波变换的图像边缘检测对铁轨图像的处理结果较好,实现了路轨两用电动车陆/轨转换的准确定位。
In order to get real-time acquisition of rail images which aims at accomplishing successful conversion of the road-rail vehicles between the site and the railroad tracks, this paper introduces a method of image processing based on multi-scale wavelet transform and adaptive threshold. It choses the way of compactly supported quadratic B-spline wavelet to make a multi-scMe edge detection of the rail image acquisition from the t^ansducer, and gets the final edge image by combining the methods of non- maximal suppression and adaptive threshold. Tt compares several classical operator of edge detection, the processing results of rail image by image edge detection based on wavelet are much better, which means the accomplishment of the exact location of the road/rail conversion of road-rail vehicle.
出处
《信息技术》
2013年第10期66-69,共4页
Information Technology
基金
上海市重点课程建设项目(1K-12-302-003)
关键词
多尺度小波变换
非极大值抑制
自适应阈值
边缘检测
multi-scale wavelet transform
non-maximal suppression
adaptive threshold
edge detection
