Tripod机器人的视觉处理应用研究

Application Research of Tripod Robot Vision Processing

针对生产线上的Tripod机器人,设计了图像采集与变换、预处理及边缘轮廓的提取、几何体形状检测及定位、几何体的颜色识别及旋转角度计算四大模块,并进行实验验证。结果表明该算法能够实现物体形状与颜色的准确识别,位置与角度的精确定位,研究成果可用于Tripod机器人的精准抓放行为。

Four modules are designed for Tripod robot on production lines lnclucdng image acquisition and transformation, image preprocessing and edge contour extraction, geometry shape detection and location, color recognition and rotation angle calculation of object. Verification is carried out with experiments. The results show the algorithm can accurately identify shapes and colors, position location and calculate rotation angle of objects. The research results can be used for precise grasping behavior of Tripod robots.