Mapping the surrounding space with the use of odometry, inertial navigation, structured light and com- puter vision

Annotation: The paper describes a relevant and challenging problem of automatic robot navigation. It describes a method for constructing the map of the surrounding space with active structured lighting and computer vision. As illumination, patterns are selected - one horizontal line, and three lines at a time, with the direction from the light source at different angles. The realized method for finding areas of illumination in frames of a video stream is based on the coordinates which allow to map time points of corresponding frames. The odometry and inertial navigation is used to estimate the distance. The software module is implemented in the programming language C++ using computer vision library OpenCV, which builds a map. The results of constructing the map frame-based video stream are given.

Keywords: automatic navigation, computer vision, active structured lighting, map of the surrounding space, opencv, c + +