基于STC目标跟踪的机器视觉测速算法

Speed Measurement Algorithm of Machine Vision Based on STC Target Tracking

为了准确快速对移动目标进行速度测量,提出一种基于机器视觉的速度测量方法。计算图像中目标区域椭圆参数对采集相机进行角度调整。使用改进的STC跟踪算法,在相邻帧图像目标跟踪的同时采用轮廓线匹配的方式进行尺度调整。最后提取每张图像中目标轮廓最小外接矩形的中心点坐标,通过差分法得到相对运动速度,并利用反光带长度与图像中目标比例关系来计算实际速度。实验表明,该方法利用机器视觉在非接触条件下得到较为准确的结果。

In order to measure the velocity of moving target accurately and quickly,a method based on machine vision is proposed in this paper.The ellipse parameters of the target area in the image are calculated to adjust the angle of the camera.Then,the improved STC tracking algorithm is used to adjust the scale by contour matching while tracking the target in adjacent frames.Finally,the Central Point Coordinates of the target contour minimum bounding rectangle in each image are extracted,and the relative motion velocity is obtained by the difference method.The actual velocity is calculated by the relation between the length of the reflective band and the target in the image.Experimental results show that this method can get more accurate results under non-contact conditions using machine vision.