Researchers at Florida Atlantic University have been counting cars to improve traffic flow in South Florida and, eventually, on all U.S. roads. There are more than 263 million registered passenger vehicles in the U.S. and more than 14 million registered vehicles in Florida alone.
Trying to make traffic move smoothly without a lot of manual intervention requires automated car counting techniques. Car counting techniques are usually tedious and cumbersome to perform, and they are not foolproof. These techniques require radar, infrared or inductive loop detectors and traffic cameras. A computer vision-based system is one alternative for car counting, but it is limited by weather conditions and natural light.
Researchers from FAU’s College of Engineering and Computer Science (COECS) wanted to find out a better way to monitor and estimate traffic flow using intelligent traffic surveillance systems. The team wanted to develop an automated car counting system that uses the infrastructure and cameras that are already in place that can perform both day and night, no matter the weather conditions.
The results of their study show that rain or shine, day or night, the system significantly outperformed any automated car counting methods currently in use. It has an average accuracy rate of more than 96 percent. This is far above the accuracy rates of the old system.
The new system, named “OverFeat Framework” has great potential in the field of traffic monitoring and could be an ideal solution for “counting cars.” OverFeat Framework is made of an effective combination of Convolution Neural Networks (CNN) and image classification and recognition techniques.
The research team, led by Hongbo Su, Ph.D., corresponding author of the study and an assistant professor in the Department of Civil, Environmental and Geomatics Engineering in the COECS, developed and implemented two algorithms for this new program: Background Subtraction Method (BSM) and Overfeat Framework using the Python language for automatic car counting.
"Understanding the physical traffic load is critical for managing traffic as well as for renovating roads or building new roads," said Su. "Counting cars is necessary in order to understand the density of cars on our roads, which ultimately helps engineers and decision makers in their planning and budgeting processes."
While developing the system, the researchers took into consideration any other factors that may affect the video cameras, like vibrations on bridges. The team studied buses (1,300 images), cars (1,300 images), taxis (1,300 images), trucks (1,568 images) and fire rescue vehicles (1,300 images) using six traffic videos that were located on the busiest roads in South Florida. The team collected footage from these cameras at different times through the whole day.
It is estimated that more than 1 million video cameras are placed along major roads like highways, freeways, motorways, expressways and arterial roads throughout the U.S. There are thousands of cameras on busy Florida roadways to help drivers with their everyday commutes.
"The best part of this new system is that you don't need any extra infrastructure because the cameras are already placed at strategic locations on our roads and highways," said Aleksandar Stevanovic, Ph.D., co-author of the study, associate professor of FAU's Department of Civil, Environmental and Geomatics Engineering, and director of the University's Laboratory for Adaptive Traffic Operations and Management. "We are utilizing videos from these cameras to accurately count cars to give us better knowledge about congestion on our roads. Then, we will share this information with traffic management specialists so that they can figure out how best to address the issues to optimize driving, provide new routes and ultimately improve traffic flow."
The team plans to work with all levels of government agencies and commercial enterprises to maximize the benefits of the system and provide a new way of “counting cars.”