Developing a Parking Monitoring System using Mask-RCNN

Abstract

In this paper, we present a technique for monitoring of parking spaces availability in unorganized parking spaces and organized parking spaces. It enables user to detect vehicles in a dedicated parking lot or in any unorganized parking space. The collected historical data helps us predict real time availability of parking at any given point in time based on these predictions we can also determine which areas can be converted to organized and legal parking spaces with help of efficient data mining techniques. Parking Monitoring System determines availability of parking space based on analysis of images obtained from video surveillance cameras. The system is developed to determine the best suitable and accessible parking location, and notify the users. We provide a system based on Mask-RCNN for analyzing the images and determining the availability of parking spaces in parking lots.