Development of a Vision-Aided Navigation Algorithm

Research toward unmanned mobile robot navigation has gained significant importance in the last decade due to its potential applications in the location based services industry. The increase in construction of large space indoor buildings (such as hospitals, malls, airports, warehouses, etc.) has made difficulty for humans to operate within such environments; especially when it involves moving long distance from one end to another or transporting heavy objects. Although mobile robot navigation solutions have been extensively applied in outdoor, it remains as a challenging task for indoor environments since the existing sensors used for navigation presents several limitations when it comes to operate in indoors. In this thesis, an indoor navigation algorithm is developed with vision camera as main sensor for the mobile robot. Using two monocular cameras (one looking forward and one looking downward), the developed algorithms make use of the salient features within the environments to estimate rotational and translational motions for real-time positioning of the mobile robot. At the same time, an algorithm based on passive beacon recognition is developed. The passive beacon is shaped arrow based signboards with different colours representing different paths. Thus, these algorithms are integrated into a designed framework, implemented on a mobile robot, for an autonomous navigation system