An Approach Using Low-Cost Drone Imagery with an Image Analysis Model to Evaluate Disaster Recoveries

Abstract

After disaster events, community leaders face the challenge of rebuilding societal infrastructure and managing the allocation of funds that can extend or reduce durations of recovery periods. Decision-makers must quickly determine how to allocate financial resources while minimizing the population distress. Conventional methods of assessing damage and evaluating relief requirements fall short of meeting the urgent recovery needs after a disaster, which could lead to negative effects on communities, such as involuntary relocation and neighborhood gentrification. This evaluates current methods and technologies and suggests a new approach using low-cost consumer drones and modern image analysis techniques to aid initial damage assessments and track recovery progress to promote dynamic appropriation. Using drone imagery allows for quick data collection and dynamic analysis, enabling multiple reviews during the disaster response and recovery phases. The study explores the potential of temporary blue tarps ("blue roofs") as a metric of recovery progress and validates the automated analysis. This research analyzes a case study of images collected during the 2020 tornado in Middle Tennessee. By providing an affordable (i.e., low-cost drones) and efficient data analysis tools (i.e., modern image analysis techniques), the goal of this research is to improve resource allocation and decision-making in post-disaster recovery efforts by government officials.