基于视觉的钢丝捆扎机C形钩姿态检测方法研究

C-hook Attitude Detection Method of Steel Wire Binding Machine Based on Vision

自主研发了一种基于线结构光的钢丝捆扎机C形钩姿态检测系统。该系统首先通过背景差分法提取特征光条并对其降噪处理,然后通过Hessian矩阵法得到精确的光条中心点坐标,最后通过相机与光平面标定结果将像平面坐标系转化为世界坐标并拟合C形钩表面方程,完成C形钩的姿态检测。最后,在以三坐标测量机为基础构建的实验平台上,对检测系统的精度进行实验验证,结果表明:算法的角度最大偏差为0.4483°,满足项目的要求。

A C-shaped hook attitude detection system for wire binding machine based on line structure light self-developed in this paper.Firstly,the system extracts the feature strips by background difference method and reduces the noise.Then,the Hessian matrix method is used to obtain the coordinates of the center point of the strip.Finally,the image plane coordinate system is converted into the world coordinates by the camera and the light plane calibration result.Fit the C-shaped hook surface equation to complete the attitude detection of the C-shaped hook.