3D Reconstruction of an Accident Site

This paper presents a method for creating a 3D-digital model of an accident site based on geophysical research practices. The model is created using photogrammetric techniques based on data from unmanned aerial vehicles. The method does not require expensive equipment or software, as all measurements can be taken using consumer-grade drones, and generally, freely available software is sufficient for calculations. The paper describes the actions of emergency responders at the incident site, but emphasizes the technical aspects of collecting data and creating models. Traditional methods require specialized equipment and skilled personnel, which can make their application challenging. Instead, we propose using miniature unmanned aerial vehicles (UAVs) to obtain photographic data for 3D-software reconstruction. However, the accuracy of this data is limited due to hardware constraints. To address this issue, we propose a new technique that involves taking more images. This allows us to compensate for errors in onboard navigation solutions and achieve accuracy comparable to full-scale measurements. We demonstrate the application of this method using a consumer-grade UAV and analyze the results using specialized software in a simulated incident. The paper also discusses the potential risks associated with using UAVs at low altitudes near roads, such as driver distraction and the possibility of collisions. Despite these risks, the technology has been successfully tested in several incidents and has been recognized as an important tool for gathering information from accident scenes.