Car Parking with Empty Slot Detection using M...

Car Parking with Empty Slot Detection using Machine Learning

Author Name : Soham Kapasi, Hetanshi Patel, Viral Patel, Shalini Shah

ABSTRACT

Parking space traffic congestion is a major issue that modern society is now dealing with, since car numbers are fast increasing without equivalent increases in parking slots. The study done here, employing Instance Segmentation techniques and Deep Learning, aids in addressing the traffic congestion problem at the bottleneck of the networks, mostly at parking spots. The model gathers all the initial available parking spots in the defined region, processes the data in real time, decides if the slots are empty or filled by any automobile, and offers information about vacant slots. The model not only discovers a free parking area for a car, but also a good parking location for a two-wheeler. With a mask identification rate of more than 92.33 percent and a border recognition rate of 98.4 percent, the proposed system is more robust. Nowadays, parking a car is a challenging task. The parking problem is aggravated by a scarcity of available parking spaces, a lack of precise information about available spaces, and the search for such unoccupied spaces. Finding a parking place not only consumes time, money, and effort, but it also makes driving more difficult. The problem worsens, causing traffic congestion, air pollution, and environmental damage. As a result, there is an urgent need for a comprehensive parking system to be implemented in order to decrease parking concerns. This research is simply an overview of a few of the parking solutions that have been proposed to manage parking concerns. This survey research addresses several ideas and implementations offered by the authors and compares them based on their efficiency, optimization, cost, and a few other variables.

Keywords—Mask R-CNN, Instance Segmentation, Vehicle Parking System, Automatic Vehicle Parking Detection, Vision Based Vehicle Parking System