基于机器视觉的栅栏自动喷漆系统设计

摘要

In view of the traditional manual painting harm to the human body, environmental pollution, waste paint and other problems, put forward a method of combining automatic painting algorithm based on machine vision servo control. The algorithm first uses the camera to collect the original fence image, then the image smoothing, threshold segmentation, skeleton extraction and processing of image and get the fence, fence image a series of 2D coordinates to complete the automatic painting system. Through online debugging shows that, this algorithm can give the coordinates by extracting the center line and center line of accurate each point in the world coordinate system to achieve the requirements of the automatic painting system. The algorithm adopts step Gauss filtering, regional segmentation, skeleton extraction, line separation, coordinate calculation and so on, finally it can create the automatic painting trajectory automatically.%针对传统人工喷漆伤害人体、污染环境、浪费油漆等问题，提出了一种基于机器视觉结合伺服控制的自动喷漆算法。该算法首先利用相机采集原始栅栏图像，然后对栅栏图像进行图像平滑、阀值分割、骨架提取等处理，进而得到栅栏图像一系列二维坐标，来使喷漆系统完成自动喷漆的方法。通过联机调试表明，该算法可以通过提取中心线后准确的给出中心线每个点在世界坐标系下的坐标，能够达到自动喷漆系统的要求。算法采用了高斯滤波、区域分割、骨架提取、线段拆分、坐标计算等步骤，最终可自动生成自动喷漆轨迹。