Establishing and Fulfilling Information Requirements for Computer Vision–Enabled Digital Twin–Based Control of Roadside Vegetation

Unattended roadside vegetation presents a notable hazard to road safety, frequently contributing to traffic accidents by blocking drivers’ sightlines and reducing visibility. This research investigates current strategies for managing roadside vegetation, evaluates their shortcomings, establishes technology-agnostic information requirements for effective vegetation control, and introduces two solutions based on computer vision–enabled digital twin technology to fulfill these requirements. The research employs a mixed-method approach, including expert interviews and a literature review, to analyze existing vegetation control processes and their shortcomings. It then defines the information requirements for digital twin implementation and proposes two computer vision–based solutions: a video-based two-dimensional (2D) approach and a mobile mapping–based three-dimensional (3D) approach. These solutions were tested through real-world case studies to validate their effectiveness in fulfilling the key requirements. The 2D approach, while affordable, provides basic assessments. In contrast, the 3D approach offers comprehensive spatial details and superior data quality for detailed analysis.