焊接机器人实时引导中光条快速提取算法研究

在结构光视觉测量系统中,激光条纹的提取精度和速率直接影响到最终测量结果的精确度。为了解决现有激光条纹提取方法难以同时达到抗干扰能力强、稳定性好和运算量小等问题,采用了一种优化Steger快速提取激光条纹中心的方法,利用梯度阈值分割出有效条纹区域,采用优化的扫描方式,用灰度重心法对条纹中心进行粗提取,然后用Hessian矩阵法对条纹中心进行精提取。对激光条纹提取原理进行了理论分析,并对软硬件设计进行了实验验证,取得了较好的实验数据。结果表明,该方法与传统方法相比,具有提取精度高、提取速率快的优点,这一结果对柔性传感器的高精度和实时性是有帮助的。

Research of rapid laser stripe extraction in real-time guiding of welding robots

In a structured light vision measurement system, the extraction accuracy and speed of laser stripes directly affect the final accuracy of measurement results. In order to reach strong anti-interference ability, good stability and low computation cost at the same time, an optimal Steger algorithm was proposed to extract center points of laser stripes. The gradient threshold segmentation was used to get the effective stripe area. The optimal scanning method and the gray centroid method were used extract the stripe center coarsely. Then, Hessian matrix method was used to extract pure stripe center. The principle of the laser stripes extraction was analyzed in theory and the hardware and software design were verified by experiments. Good experimental data was obtained. The experimental results show that this method has high extraction accuracy and fast speed compared with the traditional method. The research is helpful for high precision and real-time performance of flexible sensors.