基于像素数量统计的道岔转辙机缺口边缘检测算法

Switch Machine Gap Edge Detection Algorithm Based on Pixel Quantity Count

目的:针对青岛地铁13号线现有的道岔转辙机缺口图像检测系统存在的不足,提出了基于数量统计的边缘检测优化算法,以便提高缺口检测精度。方法:利用中值滤波将采集的缺口图像中的噪声滤除且可保留图像边缘信息;利用图像形态学处理基本算法中的腐蚀、膨胀、开运算和闭运算,进一步优化图像边缘,避免图像边缘轮廓的模糊或者毛糙,为后续处理提供优质图像样本;利用Canny算子提取缺口图像的边缘特征,结合动态阈值算法对阈值实时更新;经过基于数量统计的边缘检测优化算法,对图像进行最终的边缘位置确认,从而计算出缺口数值。结果及结论:经过试验验证,算法优化后的图像缺口准确率提高了近5%,准确率提高得较为明显。因此,基于数量统计的边缘检测优化算法能够减少因检测不准确原因而出现的道岔转辙机缺口数据报警概率。

Objective:To address the shortcomings of the current switch machine gap image detection system on Qingdao Metro Line 13,an edge detection optimization algorithm based on quantity counts is proposed to improve the accuracy of gap detection.Method:Median filtering is used to filter out the noise in the collected gap image and preserve the edge information of the image.Image morphology is used to process the corrosion,dilation,open operation,and close operation in the basic algorithm to further optimize the image edges,avoid blurring or roughness of the image edge contour,and provide high-quality image samples for subsequent processing.Canny operator is used to extract the edge features of the gap image,updating the threshold in real-time by combining dynamic threshold algorithm.Through the edge detection optimization algorithm based on quantity counts,the final edge position of the image is confirmed,thus the gap value is calculated.Result&Conclusion:After experimental verification,the accuracy of the image gap optimized by the algorithm is improved by nearly 5%,with a significant improvement in accuracy rate.Therefore,edge detection optimization algorithms based on quantity counts can reduce the probability of switch machine gap data alarms caused by inaccurate detection.