Augmented Reality Registration And Assembly State Detection For Antenna Board

In the production process of large-scale equipment,Complex wiring makes it easy to make an error with wiring installation,and the wiring quality inspection is verbose.The auxiliary assembly system in the augmented reality environment can give the operator guidance effectively and improve the yield of the equipment.Although the current augmented reality registration method based on artificial marker has been research extensively,the existing augmented reality registration method still has a great limitation in the practical industrial environment.In order to solve these problems,paper combines the requirements of state detection for wiring and wire assembly in the production of wire board.In the development environment of Unity3 D,this paper designs and implements an industrial augmented reality auxiliary assembly system by using augmented reality and image processing techniques.It focuses on solving the 3D registration algorithm in industrial scenario and the detection of the accuracy of wiring based on real-time images.The specific research work is as follows:(1)It researches and analyzes the conflict between the working condition requirements in the wiring assembly scenario and the existing 3D registration algorithm.The paper examines a variety of basic assembly states that appear in the wiring assembly scenario and designs device models that meet the requirements of augmented reality.After analyzing the principles of 3D registration of artificial markers and natural feature,the basic requirements for the conditions of the assembly scenario are used to analyze the advantages and disadvantages of the two kinds of registration method.(2)A 3D registration method based on square border marker is raised.In this paper,a planar marker with square border surrounding natural texture is designed by using the characteristics of artificial markers and natural feature in 3D registration.The edge of the marker is extracted by the image preprocessing method.The four key points of the marker are obtained by the fixed angle scan line.The ORB algorithm is used to find the key point order to solve the camera pose and achieve the 3D registration of the square border marker.A forward-backward error optical flow tracking method is introduced to track the displacement of the predictive model and the region of the marker,optimize the 3D registration efficiency and the stability of the augmented reality,and shrink the search region of the marker.Augmented reality registration experiments are designed to verify the real-time and effectiveness of the 3D registration algorithm.(3)The detection algorithm of wire assembly based on color image is researched and implemented in this paper.It detects key region of process through a marker;After RGB color space analysis is used to detect the color characteristics of the background and target in key region,a dynamic threshold algorithm is designed to separate targets and backgrounds for the color characteristics of the wire;the state of wire distribution was statistically analyzed by pixel precipitation,and a binary encoding rule was designed to describe the wire distribution information.Paper designs industrial augmented reality auxiliary assembly system by combining 3D registration and detection algorithms for wiring process.With an assembly scenario range of 70×125cm and image resolution of 1280×720p,the registration efficiency of the system can reach 24.51 ms,and the process detection efficiency can reach 26.36 ms.The speed of operation meets the requirements of augmented reality and is applicable to the wiring assembly environment of the equipment.