Research on the Algorithm for Solving the Indoor Vision Positioning Model of Mobile Robot

Abstract

In the analysis of the features and properties of camera based on the data, constructs the projection model and error model positioning calculation model, expanded beam method traditional adjustment model is only for image observations are optimized, constructed model with two or three dimensional positioning information integration RGB-D camera. The design experiments verified the accuracy of detection and matching, and combined with the depth measurement characteristics, we constructed two weight matrixes of observation information, thus improving the mathematical model of the whole location calculation.

Analysis and implementation of an artificial vision method for micropositioning based on a camera-screen system

“…En el pasado, diversos autores han estudiado en detalle diferentes métodos para la estimación de la posición utilizando una sola cámara [60][61][62][63]. Más recientemente, se han introducido sistemas de posicionamiento basados en marcadores como ArUco, Chilitags, ApriTags o ArToolKit, entre otros, para estimar cambios cuantitativos en distancias y orientaciones en muchas aplicaciones tecnológicas, como robots autónomos [64][65][66][67][68], vehículos no tripulados [69][70][71][72][73][74][75][76] o asistentes virtuales [77][78][79][80].…”