Product Blind Area Assembly Method Based on Augmented Reality and Machine Vision

基于机器视觉的增强现实盲区装配方法

¹ School of Mechanical Engineering, Northwestern Polytechnical University, Xi’an, 710072, China
² Dongguan Sanhang Civil-Military Integration Innovation Institute, Dongguan 523808, China

Received: 8 March 2018

Abstract

In the manual assembly of the blind area, the worker's line of sight is blocked, and the real-time state of the parts to be assembled cannot be seen, which has a great impact on the efficiency and accuracy of the assembly. Aiming at this problem, a blind zone assembly method based on machine vision and augmented reality(AR) is proposed. Firstly, the ellipse is used as the marker point. The object to be assembled in the blind area is indirectly tracked by the detection and positioning of the ellipse, and the AR visualization guide assembly is then performed by projection and the assembly is precisely guided using the principle of local error amplification. Finally, the blind zone assembly experiment based on machine vision and augmented reality is designed to verify the effectiveness of the method. The experimental results show that this method can significantly improve the efficiency of assembly work in blind areas and can effectively reduce the assembly error rate.

摘要

对于盲区手工装配，由于工人视线受阻，无法看到待装配零件的实时状态，对装配的效率和准确率造成了极大影响。针对这一问题，提出了一种基于机器视觉的增强现实盲区装配方法。首先将椭圆作为标志点，通过对椭圆的检测和定位间接追踪盲区待装配对象；然后通过投影的方式进行AR可视化并利用局部误差放大的原理精确引导装配；最后设计基于机器视觉的增强现实盲区装配实验，验证该方法的有效性。实验结果表明，此方法能显著提高盲区装配作业的效率，并能有效降低装配错误率。