CHEAPER TRAFFIC MONITORING APP UNDER INVENTION
By Prossy Tulinomubezi
A group of PhD researchers at Makerere University College of Computing and Information Sciences are inventing a traffic congestion monitoring app for developing cities which is cheaper compared to the prevailing technology and can help users to always predict their traffic time.
Rose Nakibuule, a lecturer at CoCIS carrying out research on traffic flow monitoring says they were motivated by traffic congestion in Uganda that makes people spend most of their time in jam instead of doing other productive work.
“Due to the chaotic traffic congestion in Uganda, we were motivated to do more research on how it can be managed from which we came up with an image based monitoring prototype that can operate in cities like Kampala, and it is a hundred (100) times cheaper than the existing technology,” Nakibuule said.
She adds that the available monitoring systems are so expensive yet they only focus on vehicles yet they are not the only causes of traffic thus want to come up with a system that can handle that kind of situation in the developing world.
“Traffic is caused by moving vehicles, animals and moving individuals yet the conventional systems available look at only vehicles. In the same way, the cost of the available monitoring systems which use high resolution cameras is very high since a single camera can cost over 1000 dollars,”
With the event of android phones which use solar energy and a dedicated server, Nakibuule believes traffic congestion will be solved.
“We are going to come up with a potable system which we can upload anywhere on a road where it can be in position to capture images on the road using a phone as a capturing device and send them to the server for processing,” Nakibuule adds.
Computer vision techniques will be used to take a sequence of images which will be used to calculate traffic flow speeds through manual calibration. “Vision algorithms can be used to distinguish between the motion of vehicles and non-vehicles. This will help determine the average speed in kilometers per hour (km/hr),” Nakibuule said.
“Cameras will be calibrated where we mark a rectangular region on the ground with a known length and width, capture it in I form of an image and discover the number of pixels the region is covering using the image which will help us to estimate the number of kilometers each image covers on the ground,” Nakibuule explained.
She says automatic calibration can also be done on an empty ground in case of roads which have clearly marked lanes.
If the system completed and working properly, the user will just need to see a map because all the areas with cameras will be marked on the map and the speed of vehicles that particular area can be identified.
Nakibuule emphasizes that the application will help in formulation of traffic plan of all cities thus coming up with a dynamic way of managing traffic.
“After collecting traffic speed for a long time on each of the marked roads, a congestion map can be done on the road to enable users know which road is congested at a particular time of the day and then come up with a dynamic way of managing traffic such that traffic rights in one road release people depending on the amount of vehicles on the other road,” Nakibuule asserts.
The project is funded by the College of Computing and Information Sciences and the team behind it includes Rose Nakibuule, Joseph Ssennyange, Innocent Komorubuga and John Quinn.