The vSLAM Algorithm for Navigation in Natural Environments

This article describes the Visual Simultaneous Localization and Mapping (vSLAMTM) algorithm, a novel algorithm for simultaneous localization and mapping (SLAM). The algorithm is visionand odometry-based, and enables low-cost navigation in cluttered and populated environments. The algorithm creates visual landmarks that are highly distinctive and that can be reliably detected, virtually eliminating the data association problem present in other landmark schemes. No initial map is required, and dynamic changes in the environment, such as lighting changes, moving objects, and/or people are gracefully handled by the algorithm. Typically, vSLAM recovers quickly from dramatic disturbances, such as “kidnapping”.