Institute of Transportation, MOTC

　　The acquired freeway on-ramp area traffic data includes traffic volume, vehicle types, average speed and occupancy of each lane in every pre-divide zone. Eight sites were selected to collect the above mentioned data based on the criteria of mainline number of lanes, on-ramp number of lanes, and merging area length.

　　The data collection and analysis were conducted by an Image Process Video Detection (IPVD) method in Autoscope. Video cameras were mounted on selected light poles to record the images of dynamic traffic states. The tapes were then analyzed through an image processing system in a laboratory. Traffic characteristic data files were finally converted into the designated format for the purpose of model development. With the data collection and analysis method used in this study, the required field man-power and time could be greatly reduced compared to conventional methods. Moreover, due to its low interference with the traffic operations, the data collected is more reliable. The experience gained in the study is also a valuable asset for future applications.