This repository contains a brief description about the procedure to reproduce the comparison tables presented in the survey article "Mapping the Landscape of SLAM Research: A Review." available at IEEE Latin America Transactions.
Kitchenham and Charters proposed that Systematic Mapping Studies (also known as Scoping Studies) are designed to provide a wide overview of a research area, to establish if research evidence exists on a topic and provide an indication of the quantity of the evidence. Following Kitchenham and Charters ideas, a preliminary search stage was first conducted to assess the volume of relevant studies.
As a result, several SLAM surveys and articles were found, and the following inclusion/exclusion rules were defined:
- Only articles written in Spanish or English were considered.
- Probabilistic SLAM approaches were not included since these kind of systems have already been extensively studied and reviewed.
- The search stage was carried out with Google Scholar's search engine. Conference proceedings, journal papers, and unpublished but relevant articles were included.
- Given the significant number of papers presenting results on the EuRoC, KITTI, and 7-Scenes datasets, we focused our comparative analysis on the performance demonstrated by systems on these datasets.
A table for the most popular SLAM datasets was first drawn. It allows the reader to compare different datasets in terms of length, amount of sequences, among other properties.
Performance comparison Tables were also drawn. However, in this case, some rules were established: Only works who reported objective (i.e. numeric) results were considered to draw comparison tables. Those with qualitative and graphical results were excluded. Simulations were not conducted; instead, the tables were populated with data directly extracted from the original publications of the respective systems.