Stud Pose Detection Based on Photometric Stereo and Lightweight YOLOv4

作     者：Xuan Zhang Guohui Wang 

作者机构：School of Opto-Electronic EngineeringXi’an Technological UniversityXi’an 710021China 

出 版 物：《Journal of Artificial Intelligence and Technology》 (人工智能技术学报（英文）)

年 卷 期：2022年第2卷第1期

页      面：32-37页

学科分类：08[工学] 081101[工学-控制理论与控制工程] 0811[工学-控制科学与工程] 081102[工学-检测技术与自动化装置] 

基　　金：The work is partly supported by the Natural Science Basic Research Plan in Shaanxi Province of China(No.2016JM6041) 

主　　题：Stud pose photometric stereo neural network machine vision 

摘      要：There are hundreds of welded studs in a *** posture of a welded stud determines the quality of the body assembly,thus affecting the safety of *** is crucial to detect the posture of the welded *** the lack of accurate method in detecting the position of welded studs,this paper aims to detect the weld stud’s pose based on photometric stereo and neural ***,a machine vision-based stud dataset collection system is built to achieve the stud dataset labelling ***,photometric stereo algorithm is applied to estimate the stud normal map which as input is fed to neural ***,we improve a lightweight YOLOv4 neural network which is applied to achieve the detection of stud position,thus overcoming the shortcomings of traditional testing *** research and experimental results show that the stud pose detection system designed achieves rapid detection and high accuracy positioning of the *** research provides the foundation combining the photometric stereo and deep learning for object detection in industrial production.