Three-Dimensional Location and Mapping Analysis in Mobile Robotics Based on Visual SLAM Methods

Abstract

One of the essential tasks required from a mobile robot is the autonomous and safe navigation of its working environment. However, in many cases, a model of the environment or map is not available to execute this task. Indeed, navigation requires a permanent estimation of the location for a map, which is not available for unknown environments. In such a scenario, the robot must have extended capabilities to solve, concurrently, the problems of localization and mapping. The simultaneous solution of these two problems is known as SLAM (simultaneous localization and mapping) and is a complex problem, not yet fully solved by the scientific community. This is due to the fact that localization requires a map that is not yet available since it is still under construction. In turn, the elaboration of a map requires the estimation of the robot’s location. This is the reason why SLAM has been categorized as similar to the chicken and egg problem. In the case of a robot facing an unknown environment, it would be something like what to solve first, localization or mapping? The answer to this question is that the robot will have to solve both problems at the same time. This article presents a study of some of the most representative open source visual SLAM (vSLAM) methods, beginning from an analysis of their characteristics and presenting criteria selection for an experimental design that allows contrasting their advantages and disadvantages. Two of the most representative algorithms for solving vSLAM were considered (RTAB-Map and ORB-SLAM2). The experiments were validated with a robotic system designed for this purpose, which is fully compatible with ROS (robot operating system).

A review of visual SLAM for robotics: evolution, properties, and future applications

“…V-SLAM plays a significant role as a transformative topic within the robotics industry and research ( Khoyani and Amini, 2023 ; Acosta-Amaya et al, 2023 ). The progress in this field can be attributed to tools such as machine learning, computer vision, deep learning, and state-of-the-art sensor technologies, which have collectively simplified and enhanced its strategy in real-life applications ( Beghdadi and Mallem, 2022 ; Duan et al, 2019 ).…”

“…However, versions 2 and 3 include additional steps such as bundle adjustment welding and map merging ( Mur-Artal and Tardós, 2017a ; Zang et al, 2023 ). The last stage is preparing the output, focusing on creating the final map that includes essential information such as graphs, lines, point mapping, and 2D and 3D maps for use in the SLAM process ( Acosta-Amaya et al, 2023 ). Figure 5 parts 4, 5, and 6 give a detailed observation about the methods of ORB-SLAM 1, 2, and 3 versions, respectively, showcasing their features and functionalities for a better understanding.…”

“…This choice comes from the challenge of figuring out how far things are in areas where there is not much texture inside images. As a result, LSD-SLAM goes beyond what DTAM can do by concentrating on places with strong changes in brightness and ignoring areas with very little texture ( Acosta-Amaya et al, 2023 ; Khoyani and Amini, 2023 ).…”

“…RTAB-Map SLAM, which stands for real-time appearance-based mapping, is a visual SLAM technique that works with RGB-D and stereo cameras ( Ragot et al, 2019 ). It is a versatile algorithm that can handle 2D and 3D mapping tasks depending on the sensor and data that are given ( Peter et al, 2023 ; Acosta-Amaya et al, 2023 ). It integrates RGB-D and stereo data for 3D mapping, enabling the detection of static and dynamic 3D objects in the robot’s environment ( Ragot et al, 2019 ).…”

“…It gathers real-time data about the surroundings, creating both a geometric and a visual model r13 (accessed on 14 November 2023). In addition, the process includes the implementation of bundle adjustments (BAs) to improve the precision of the generated map before it is moved to the final stage ( Acosta-Amaya et al, 2023 ). BA is a tool that simultaneously refines the parameters essential for estimating and reconstructing the location of observed points in available images.…”