Off-Road Detection Analysis for Autonomous Ground Vehicles: A Review.

Area of Science:

• Robotics
• Computer Vision
• Artificial Intelligence

Background:

• Autonomous ground vehicles (AGVs) require robust detection capabilities for safe navigation.
• Off-road environments present unique challenges for AGV perception compared to on-road settings.
• Key elements for AGV navigation include identifying drivable pathways and surrounding obstacles.

Purpose of the Study:

• To provide a comprehensive survey of recent advancements in AGV detection methods specifically for off-road environments.
• To systematically categorize existing literature based on the type of detection (drivable ground, positive obstacles, negative obstacles).
• To analyze detection methods based on sensor technology (single vs. multiple sensors) and data analysis techniques.

Main Methods:

• Literature review and categorization of AGV detection methods.
• Classification of methods based on target elements: drivable ground, positive obstacles, and negative obstacles.
• Analysis of detection approaches based on sensor fusion and data processing techniques.

Main Results:

• The survey divides off-road AGV detection literature into drivable ground, positive obstacles, and negative obstacles.
• Methods are further classified by sensor type (single/multiple) and data analysis approaches.
• Critical findings, current challenges, and future research directions in off-road AGV detection are discussed.

Conclusions:

• This survey offers a structured overview of the state-of-the-art in off-road AGV detection.
• It identifies key technological trends, persistent challenges, and promising avenues for future research.
• The work aims to assist researchers in navigating the literature and identifying related studies in the field.