VR for Situational Awareness in Real-Time Orchard Architecture Assessment

Area of Science:

• Robotics and Automation
• Computer Vision
• Virtual Reality

Background:

• Teleoperation in agriculture demands high situational awareness for tasks like orchard management.
• Complex orchard structures challenge remote visualization during architectural scouting.
• Existing systems struggle with real-time data rendering for effective remote operation.

Purpose of the Study:

• To develop an adaptive streaming and rendering pipeline for real-time point cloud visualization in VR teleoperation systems.
• To enhance architectural scouting and decision-making in precision agriculture.
• To improve the operator's Quality of Experience (QoE) in remote agricultural environments.

Main Methods:

• Implemented a pipeline with selective streaming, a Unity Engine point cloud parser, and adaptive Level-of-Detail rendering.
• Utilized dynamically scaled and smoothed polygons for efficient rendering.
• Incorporated pseudo-coloring via LiDAR reflectivity for enhanced material and geometry distinction.

Main Results:

• Achieved 10.2-19.4% improvement in runtime performance compared to existing methods.
• Demonstrated framerate enhancements of up to 112% through selective streaming.
• Confirmed visual continuity and preservation of geometric features for scouting operations.

Conclusions:

• The proposed VR-based teleoperation system is viable for precision agriculture.
• The pipeline effectively enhances remote perception and decision-making in complex orchard environments.
• Quality-of-Service parameters critically influence operator Quality of Experience in remote sensing.