Vehicle Extraction from Aerial Images Using Voting Process and Frame Matching

In today´s motorized society, traffic congestion and illegal parking are serious problems in cities. We must collect traffic information to solve these problems. Fixed cameras or sensors set up on roads have been used so far. However, approaches using such cameras have many disadvantages. For example, these cameras cannot be used in the event of accident, and it is difficult to collect information for a wide area at one time. On the other hand, a survey of areas by many people requires much time and money. There is a real need to simplify these approaches. In our study, we extract vehicles from a wide area of roads using aerial images, and we propose a method for helping to obtain traffic information.