Embedded Automated Vision for Double Parking Identification System

Abstract

The aim of this work is to assist the city administration issue which involve the traffic flow disruption in an urban area. One of the causes of traffic flow disruption is double parking; thus, in this work, an automated double parking identification and alert system was developed using embedded vision system and internet of things. A camera was utilized to acquire the image of a parking area, and the image was processed using Beaglebone Black processor. A computer vision algorithm was developed to process the image using background subtraction, region of interest identification, and color analysis. When a double parked vehicle is detected, the data was sent into the cloud automatically to alert the city administrator for further action. The developed system achieved 91% accuracy in detecting the traffic violation of double parking.