Visual SLAM technology based on weakly supervised semantic segmentation in dynamic environment

A visual simultaneous localization and mapping (vSLAM) system in a dynamic environment are affected by the wrong associated data caused by the moving targets. Combining semantic segmentation information to remove dynamic feature points is an effective method to improve the pose estimation accuracy of the vSLAM system. However, the existing semantic vSLAM usually adopts the fully supervised methods to segment the dynamic scenes. The accuracy of supervised methods relies on a large number of training data sets with annotation information, which limits the application of SLAM system. To address this issue, a visual semantic SLAM system (vsSLAM) that applies weakly supervised semantic segmentation to dynamic scenes is proposed to broaden the application range of the system. Firstly, the system extracts the feature points of input image and checks the movement consistency, and then segments the dynamic target with the weakly supervised methods. Secondly, the semantic segmentation results are used to remove the dynamic feature points in the image. Finally, the system uses stable feature points for pose estimation. Furthermore, the Automatic Color Equalization algorithm is adopted to pre-process the input image to improve the accuracy of weakly supervised semantic segmentation. Experiments were performed on the public TUM dataset and lab environment data. The results show that the accuracy of the vsSLAM system based on our method is better than the traditional ORB-SLAM2 system, and also higher than the SLAM system of the weakly supervised network DSRG. The accuracy is close to the fully supervised semantic SLAM system.