In gap acceptance studies, it is common to obtain data of time gaps from traffic interactions by playing back the video recordings of the traffic site and measuring the headways between vehicles in the traffic streams. This method of data extraction can be rather tedious and subject to substantial human errors. This paper proposes an automated process of obtaining gap information from videotapes by using data output from the ARRB video-image processing software CAMDAS. To assess the reliability of the proposed method, the vehicle counts and the distributions of headways and accepted gaps extracted from CAMDAS output were compared with the corresponding gaps that were extracted manually. Favourable results from the comparisons indicate that CAMDAS may be used in an automated process to obtain the accepted gap distributions. (a).
Samenvatting