Road Traffic Density Surveillance System using IR Sensor, Image Processing and Raspberry Pi: A Comparative Analysis

In our present time, the use of vehicles has become widespread which has given rise to the problem of traffic congestion a common sight in day-to-day life. And with the increased sales of vehicles in recent years the problem is almost considered as a known factor that gives rise to noise and air pollution in coming years. This imminent problem gives validation to the use of de-congestion methods, some of which are more efficient traffic signs and constant traffic monitoring to prevent a standstill event that is known to devolve into traffic jams. Traffic monitoring can be done with the help of Frame comparison as well as Edge Detection in tandem with IR sensors that give a rough estimate of how far the traffic congestion goes. In this paper we have focused on the merging of three different approaches viz. the IR Sensor in passive mode, Real-time Background Subtraction in a semi-active mode, and Real-time Canny Image Processing in an active mode in an efficient manner in a prototype and analyzed their performance using Raspberry pi for selection of a preferred system. The results obtained reduce traffic congestion using the dynamic times allocated to the higher traffic lanes until an average time, which is assumed with lower traffic lanes. This method is inexpensive and easily implemented with Raspberry Pi's help to run it arguably efficiently on medium-traffic areas.