Time headway between vehicles is considered as an important microscopic traffic flow parameter that affects the safety and capacity of highway facilities. This research work intends to provide a field study of vehicle time headway distribution on Petra multilane Highway that connects Irbid and Ramtha cities in Jordan.  The main objective of this study is to integrate the utilization of computer vision and Artificial Intelligence (AI) to extract time-headway data, and to investigate the suitability of the negative exponential distribution for random headway state and the normal distribution for constant headway state. Time headway data were videotaped for moving traffic over two different periods, one having medium traffic volume and the other representing the rush hour.   Five-hundreds observations of time headway were extracted through an AI Python code developed specifically for this task and vehicle’s image detection. 50% of the observations were random headways state and the other 250 observations were constant headways state.   The developed regression analysis model for the extracted time headway data versus their associated frequencies shows tendency toward negative exponential distribution (R²=0.98) for random headway state, while it shows a normal distribution relationship (R²=0.99) with Chi-square test at a level of significance of 0.01. This indicates the successful integration between computer vision and AI in investigating suitable models for headway distributions. This finding will open the door for the usage of AI and computer vision in numerous applications of traffic and transportation engineering disciplines.