摘要
针对铁轨图像在采集过程中经常受到不同程度的噪声影响,传统边缘检测方法难以检测并提取准确的铁轨边缘。文中通过分析灰度数学形态学抗噪图像边缘检测的常用方法,提出基于不同几何形状及大小的双结构元素抗噪数学形态学铁轨图像边缘检测算法。并应用这种算法对掺杂有噪声的铁轨弯道图像边缘进行铁轨边缘检测,实验结果表明:该方法具有比传统经典边缘检测方法更好的铁轨弯道边缘检测及提取效果。
Track images were often subject to different levels of noise in the collection process in outdoor, it became difficult to detect and recognize the true edge of track by the traditional edge detection algorithm. By analyzing the commonly used gray-scale mathematical morphology edge detection algorithms of anti-noise, this paper proposed mathematical morphology anti-noise track edge detection algorithm based on different geometry and size of the dual-structure element. And by which the edges of the track comers image polluted by noise were detected. Experiments showed that this approach had better edge detection and recognization than the classic edge detection algorithm to the track comers image with noise doped.
出处
《铁路计算机应用》
2011年第5期28-31,共4页
Railway Computer Application
基金
高等学校博士学科点专项科研基金项目(20060732002)
甘肃省自然科学基金项目(096RJZA084)
甘肃省教育厅研究生导师科研计划项目(0814-4)
关键词
灰度数学形态学
边缘检测
结构元素
轨道边缘检测
gray-scale mathematical morphology
edge detection
structural elements
track edge
