The STDyn-SLAM: A stereo vision and semantic segmentation approach for SLAM in dynamic outdoor environments

TL;DR

STDyn-SLAM is a real-time stereo vision and semantic segmentation-based SLAM system designed for dynamic outdoor environments, effectively detecting and avoiding moving objects to improve mapping and localization accuracy.

Contribution

It introduces a novel SLAM approach combining stereo vision, neural network-based object detection, and segmentation to handle dynamic scenes in outdoor environments.

Findings

Achieves real-time performance in dynamic outdoor scenes

Outperforms state-of-the-art SLAM methods in dynamic environments

Effectively detects and segments moving objects to improve localization

Abstract

Commonly, SLAM algorithms are focused on a static environment, however, there are several scenes where dynamic objects are present. This work presents the STDyn-SLAM an image feature-based SLAM system working on dynamic environments using a series of sub-systems, like optic flow, orb features extraction, visual odometry, and convolutional neural networks to discern moving objects in the scene. The neural network is used to support object detection and segmentation to avoid erroneous maps and wrong system localization. The STDyn-SLAM employs a stereo pair and is developed for outdoor environments. Moreover, the processing time of the proposed system is fast enough to run in real-time as it was demonstrated through the experiments given in real dynamic outdoor environments. Further, we compare our SLAM with state-of-the-art methods achieving promising results.