| Summary: | Korea has seen that the installation of CCTVs is becoming more common to control over-speeding vehicles, track fugitive vehicles, analyze accidents, and prevent crime. As CCTV becomes cheaper and its need is increasing to maintain public order and security, it is now widely distributed. When estimating a vehicle's speed through CCTV images, a vehicle's driving distance is measured by using road lanes or surrounding structures. The speed is then calculated by analyzing travel time. Furthermore, if there is no lane or structure in the section where vehicle speed is to be estimated, it is difficult to measure driving distance. Thus, estimating speed becomes difficult. Likewise, speed is estimated by using the cross-ratio method. However, when estimating vehicle speed by using cross-ratio in CCTV images, lens distortion and curve-section errors occur. Since it is difficult to use the cross-ratio method when the front and rear wheels are not visible at the same time in the video, this study is proposing a new method. The method of estimating vehicle speed as presented in this study is to estimate the speed of an accident vehicle by comparing the location and time of the accident vehicle and the test vehicle in the CCTV image after filming the test vehicle with CCTV at the accident site, while driving the test vehicle at the same trajectory as the accident vehicle at a constant speed. This method is expected to be used in actual events.
|
|---|