Control of a transport robot in extreme weather based on computer vision systems

Abstract

Abstract This article presents a new way to use special computer vision techniques to aim with assisting in controlling a transport robot in a dynamic environment under exceptional and difficult environmental conditions. An analysis and development of algorithm for obstacle detection in the robot’s environment proposed based on data from an RGB-D video camera using computer vision methods. Contour analysis was the base method to detecting objects featured fragment taking into account difficult vision conditions. Based on open-source library (Open CV), we adopted methods program implementation which confirmed its applicability to detect objects in mobile robot environment.