基于机器视觉的工程机械行驶速度的研究

工程机械行驶速度是预测机器的牵引力、滑转损失、牵引效率和滑转率等牵引性能参数时所需要的基本变量，传统的接触式测量方法由于行驶路面条件等因素的影响，所测量的速度值并非是工程机械的实际行驶速度，利用工程机械行驶路面的纹理随机性，提出了基于机器视觉技术的工程机械行驶速度测量方法，该方法通过安装在车体上的CCD摄像机，在连续成像的条件下，通过匹配跟踪得到路面同一纹理特征在相邻两帧图像中的位置，利用摄像机内外参数和成像帧频实时计算得到工程机械的瞬时行驶速度，通过采集路面纹理图像进行模拟实验验证了该方法的可行性。

Machine-vision-based study on driving speed of construction machinery

The driving speed of construction machinery is one of the basic variables to predict such traction property parameters as traction force, slipping loss, traction efficiency and slipping rate. Owing to driving surface conditions, the measured speeds from traditional contact measurement methods are not equivalent to actual driving speeds. Based on the textural randomness of driving surfaces, a machine- vision-based technology is first proposed for driving speed measurement. By matched-tracking the consecutive images via the CCD camera installed on vehicle body, the position in two neighboring frames of the same textural property is then obtained. With interior and exterior parameters and image frame frequencies, the real-time instantaneous driving speeds are calculated. Finally, the feasibility of the proposed method is verified via experimental simulation by capturing surface texture images.