Visual SLAM Research Based On ORB-SLAM2

Vision-based simultaneous localization and mapping(VSLAM)is currently an important research direction in the fields of 3D reconstruction and computer vision,and is the core technology in the fields of autonomous driving and augmented reality(AR).At the same time,with the popularization of mobile devices and the improvement of computing power,mobile augmented reality has shown high practical value.Aiming at the current shortcomings in the SLAM,this thesis improves the algorithm and engineering of the core modules of SLAM,proposes an efficient and robust visual SLAM,and implements augmented reality applications based on SLAM on mobile platforms.The main work of this thesis can be summarized as follows.Improve the accuracy and robustness of SLAM using deep learningbased feature and traditional descriptor that draw on deep learning training techniques.This thesis takes ORB-SLAM2,a classical framework in the field of visual SLAM,as the baseline,improves the system implementation of ORB-SLAM2 and uses HFNet feature and Box Average Difference(BAD)descriptor to enhance the performance of the SLAM.The results tested on publicly available datasets show that the improved system can significantly improve the accuracy and robustness of the system while ensuring real-time performance,demonstrating the superiority of deep learning-based networks and their advanced techniques over traditional methods.A GPU-based nonlinear optimization framework g2o-GPU is proposed,and the nonlinear optimization module of ORB-SLAM2 is implemented using g2o-GPU.Nonlinear optimization has the largest time overhead among all modules of SLAM.This thesis design and implement the nonlinear optimization framework g2o-GPU on GPU using the powerful parallel computing capability of GPU,and deploy it to ORBSLAM.The experimental results show that the improved system can reduce the time overhead of the system by 3 to 5 times.Realize mobile augmented reality application based on ORB-SLAM2 and Unity rendering engine.ORB-SLAM2 realizes scene reconstruction and camera localization,and Unity efficiently renders real scenes,bringing users a cool experience.At the same time,based on the Tomcat server,the application is extended to an AR application that can interact with multiple people in real time,which has important application value.