| Popis: |
This paper introduces definition of accident (restricted to pedestrian falls) type through the analysis of risk factors on pedestrian accident cases. So far, there have been various developments in services to auto-recognize accidents to prevent fatal accidents of pedestrians. However, there are few existing services for this. The best reason for this is that the analysis of the accident behavior that we want to get from the results has not been done properly. Through analysis of actual accident cases, we present the risk factors that directly affect human injury, and the types of accidents are defined based on them. In order to reproduce and analyze the defined accident types using experts who had performed fall movements for many years, we have developed our own wearable devices with built-in acceleration sensors that can recognize pedestrian behavior. And, the collected data from these sensors is transmitted to smart-phone by using the bluetooth communication. Finally, the data by accident type collected and data by accident type in other papers are analyzed from the perspective of risk factors, explaining that the risk factor result values of existing studies are similar to normal data. |
