Perception technology for conversion of off-road vehicles for the purposes of unmanned missions

Abstract

Autonomous ground vehicles (AGV) have great potential for a wide range of applications, both in the civilian and military fields. Nowadays there is an interest in converting existing off-road vehicles into remotely controlled platforms with autonomous mode due to several benefits, including reducing the risk to human life, increasing efficiency and accuracy, and allowing the vehicles to operate in hazardous zones. Environmental perception technology plays a critical role in enabling the safe and effective operation during unmanned missions. This technology involves sensors, cameras, and other devices to gather information about the environment and provide the unmanned ground platform (UGV) with a perception of its surroundings. In recent years, there has been significant progress in the development of environmental perception systems, including the use of advanced sensors, machine learning algorithms, and other innovations have become a focus of research and development for many countries. This paper describes a combination of commercially available vision sensors, laser scanners and navigation modules for comprehensive understanding of operational environment, orientation, objects recognition during autonomous mode. Typical methods for vision and lidar-based obstacle detection and object classification for unmanned vehicles are described. The aim of the work was to examine in real environment the performance of a perception system that was configured using a daylight-thermal observation and stereo cameras, lidar sensor, GNSS module, radio links along with computing units. This system was evaluated in terms of performance of different sensors considering implementation for all terrain vehicle as subsystem for unmanned mode.