提出了一种基于视觉测量的挖掘机工作装置姿态测量方法。使用工业相机获取工作装置侧视图像,采用鞍点检测方法快速捕捉工作臂上的人工靶标;将靶标间的固定几何尺寸关系作为约束条件筛选出靶标中的鞍点并计算相应工作臂的姿态角;通过预判靶标的运动范围以缩小图像检测区域,提高算法处理速度。试验表明,与挖掘机上原有的拉线传感器测量系统相比较,动臂和斗杆姿态角动态测量偏差分别小于1°和2°,处理每帧工作装置运动图像平均用时在100 ms以内,验证了该方法对挖掘机工作装置姿态测量的可行性。
An attitude measuring method for excavator' s manipulator based on vision measurement was proposed. Using the industrial camera to grab the perspective images of the manipulator, a saddle point detection was adopted to detect the man-made targets which was fixed on every arm of the excavator. Then the fixed geometric dimensioning between two targets was used as a constraint condition to select the saddle point in targets, and the attitude angle of the corresponding arm can be derived from the saddle point coordinates. The image detection area was reduced through anticipation of target movement and the algorithm processing speed was improved. Experiment results showed that compared with the original displacement sensor system installed on the excavator, the measured deviations of dynamic attitude angle for boom and arm were less than 1 degree and 2 degrees, respectively. The average processing time of this method for each manipulator movement frame was within 100 ms. And the feasibility of this method for measuring excavator manipulator' s attitude was verified.