An indoor mapping and localization method based on Digital Landmark Map (DLM) through MI-CSURF is proposed to meet the problem of low localization speed caused by complex mapping algorithm of existing mobile robots in large indoor environments. Firstly, according to the characteristics of the environment, construct CSURF features that are stable to ambient light and viewing angle changes as the characteristics representation of DLM. Secondly, the redundant frames under the global color measurement of the image sequence are removed on the basis of color mutual information entropy (Colored MI), and final DLM is further extracted base on the image feature similarity index. Finally, the camera pose is calculated by the method of re-projection error. The loop closure and position optimization are carried out by using the DLM, to improve the accuracy and real-time performance of the localization system. Experiments show that the indoor DLM data mapped by this algorithm has a small amount of data, and the average relative error of localization in the actual scene is about 0.79 m.
Mapping and Localization Based on Indoor Digital Landmark Map Through MI-CSURF
Lect. Notes Electrical Eng.
2021-10-30
12 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
ALIMC: Activity Landmark-Based Indoor Mapping via Crowdsourcing
Online Contents | 2015
|LANDMARK-LESS SIMULTANEOUS LOCALIZATION AND MAPPING TECHNICAL FIELD
Europäisches Patentamt | 2022
|LANDMARK-LESS SIMULTANEOUS LOCALIZATION AND MAPPING TECHNICAL FIELD
Europäisches Patentamt | 2023
|