Research On Robot Shoe Gluing Technology Of Monocular 3D Vision Based On Line Structure Light

With the continuous development of automated processing technology,the way of replacing artificial production by robots has gradually penetrated from the heavy industry with heavy labor to the labor-intensive light industry.Although robots work more efficiently than artificial ones,how to overcome the problems of diversification,flexibility and large batch size in light industrial production has become a big problem in liberating labor from the current factory.In the past ten years,with the development of non-contact measurement technology represented by machine vision,robot processing with machine vision technology has become an indispensable part of modern production and processing.Machine vision technology enables efficient and accurate 3D measurement of individualized object,providing control points for subsequent path tracking for robotic servo control.This paper proposes a monocular 3D vision measurement based on line structure light and an industrial robot vision control scheme for visual 3D measurement and industrial robot processing in the flexible production line of the shoe industry.Monocular 3D reconstruction and robot vision servo control Processing and research,mainly for the following aspects of research work.(1)For the established application scenarios,scope of use and accuracy requirements,combined with the line structure optical monocular vision measurement module and industrial robot motion module,a complete set of monocular 3D vision industrial robot servo control processing system was designed and built.The system includes the shoe upper visual measurement spray station and the special-shaped sole visual measurement spray station,which can efficiently complete the three-dimensional reconstruction and trajectory extraction of the tested object,and complete the autonomous control of the industrial robot processing.(2)Improved the traditional monocular vision model based on the first structured light.Applying scheimpflug principle to improve the model imaging principle,the new imaging model obtains a clear depth of field in the optical plane of the online structure in the full field of view,and proposes the corresponding visual calibration method and the hand-eye calibration algorithm combined with industrial robot.(3)A set of versatile line structure light monocular vision algorithm was designed.The corresponding image processing flow method and 3D point cloud processing algorithm are proposed to improve the precision of the processing result.A method based on image initialization 3D point cloud is proposed,and then point cloud processing and target feature extraction are completed.The software and hardware platform is built to verify the effectiveness and practicability of the algorithm.