基于双目图像匹配的车载测速测向方法研究

Research on Vehicle-Mounted Speed and Direction Detection Method Based on Binocular Image Matching

为实现车辆对自身车速与方向的准确检测,提出了一种根据2个并列车载成像设备拍摄的路面信息图像完成速度与方向检测的方法。通过精确控制2个成像设备间的拍摄时间间隔,获得相互重叠的路面信息图像,利用图像匹配技术计算图像的相对运动距离与方向,并结合标定参数和间隔时间,计算获得准确的车速与方向。试验结果表明,该方法具有较高的可信度和精度。

To detect vehicle speed and direction accurately, the paper proposed a method based on the road surtace information images captured by two parallel vehicle-mounted imaging devices, By controlling imaging interval of two imaging device accurately, the overlapping road surface information images can be obtained. Using image matching technique to calculate the relative motion distance and direction, and in combination of calibration parameters and interval time, the accurate vehicle speed and direction can be calculated. Test results prove that the proposed method has high credibility and accuracy.