The unmanned transportation system will play an important role for lager scale farming. The control system of a transportation vehicle, traveling autonomously on farm road, has been developed based on image processing. The objective of this research is detecting the boundary lines between farm road and farm field. The XYZ image data was applied for boundary line detection between farm road and field. The method consisted of “Whole image processing” which was useful for line detection method before the vehicle running, in order to know its last position and orientation, and “Limited image processing” which was useful for the boundary line detection during the vehicle running based on the previous boundary line parameters. As the result of running test, limited image processing could detect the boundary lines at 10.1Hz and the prototype could run on a 100m straight farm road.