LFD-SLAM: improved point line combination feature algorithm based on ORB-SLAM2

With the progression of computer and automation technologies, alongside advancements in the shipbuilding industry, the automation level in ship manufacturing has been steadily increasing. This paper addresses the issue of unclear textures and uneven lighting in the ship bottom environment, which prevent the ORB-SLAM2 algorithm from extracting an adequate number of feature points. To address this issue, this paper proposes a method that combines point and line features based on the LSD algorithm. When the number of feature points is insufficient, line features are extracted to replace point features, and feature matching computations are performed exclusively in the mapping thread. Finally, a simulated model of the real shipyard environment and the robot was created using Gazebo software, and comparative validation demonstrated that the improved algorithm proposed in this paper achieves higher accuracy than ORB-SLAM2.