基于RGB颜色模型的轨道轮廓识别算法

摘要

在非接触轨距测量系统中,提出一种基于RGB模型轨道轮廓的识别算法.该算法利用光源颜色相对固定的特点,通过提取RGB图像中的R颜色特征分量,并结合图像差影算法和细化算法,准确提取出钢轨断面轮廓.实验结果表明:该算法具有抗噪声能力强、计算速度快、准确性高等特点.%In the non-contact gauge measuring syslem,a gauge recognition algorithm of track contour based on RGB model is proposed. Using the characteristics that light source color is relatively fixed,the algorithm extracts the R color features of RGB image, and combined with image difference algorithm and thinning algorithm, rail section contour is accurately extracted. Experimental results show that the algorithm has characteristics of 9trong anti-noise ability .fast calculation rate,high accuracy.