River navigation system using Autonomous Surface Vessel

Abstract

This paper presents a navigation system for an Autonomous Surface Vessel (ASV) capable of maneuvering in riverine environments. A real time vision-based navigation system is used for ASV to travel along the river. The ASV is equipped with propeller and rudder to control the speed and heading direction, respectively. Continuous stream of images are captured by camera that mounted on ASV and sent to the base station over network in real time. An image segmentation algorithm based on Hough Transform technique is implemented for waterline detection. The purpose of using Hough Transform is to perform visual based navigation that track along with the river. Besides, a high accuracy GPS receiver is installed on the ASV to provide the latitude and longitude coordinates. Data logger system helps to save the data and important information for further analysis and processing. The optical flow algorithm is implemented to detect and avoid obstacle. The ASV's navigation, control and task specific vision have been evaluated through experiments with results presented to demonstrate its capabilities.