Intelligent Yellow Interval Actuation Using Computer Vision Automated Traffic Parameter Sensing and Extraction

One of the challenges of road safety is the uncertainty in deciding whether to continue crossing or to stop at an intersection during yellow or amber light signal. The issue arises from the dilemma zone when approaching a yellow light, often caused by insufficient yellow interval duration when using the road speed limit derived from the 85
^th
-percentile of the recorded speeds of vehicles crossing in the kinematic equation computation, whereas the actual speed of crossing vehicles is often greater than the speed limit. This work demonstrated the ability of the developed Computer Vision based algorithm to determine the kinematic equation's optimum yellow interval duration by using object tracking to derive the approach speed of the test vehicles. It was concluded that the computer vision algorithm and the ground truth from the vehicle's speedometer are functionally congruent and there was no statistically significant difference between the outputs of the two methods with the variance attributed to normal distribution error.