Scribd
Upload
28 views11 pages

Smart Parking System

Uploaded by

A0554
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
28 views11 pages

Smart Parking System

Uploaded by

A0554
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 11

Smart Parking System

Group No-5
21011A0505:Arthi Chebrolu
21011A0516:Kadaru Varsha
21011A0554:Venkata Shravya Karyampudi
Table of Contents
01. Introduction 06. Mask R-CNN Algorithm

02. Abstract 07. System WorkFlow

03. Motivation 08. Conclusion

04. Existing Model 09. References

05. Proposed Model


INTRODUCTION

Smart Parking Systems use technology to efficiently manage


parking spaces. The system detects available parking spots
and informs drivers in real time. This project uses Machine
Learning (ML) and Computer Vision to detect cars and empty
spaces in parking lots. It integrates image processing, object
detection, and real-time data visualization to automate
parking management.
Abstract

This project aims to develop a smart parking spot detection


system using Mask R-CNN, an advanced object detection
algorithm known for its precision in detecting objects at the
pixel level. By processing video feeds from surveillance
cameras, the system accurately distinguishes between
occupied and free parking spaces in real-time. The system
integrates with a user-friendly React.js interface, offering real-
time updates. Designed for scalability and easy deployment, it
reduces time spent searching for parking, mitigates traffic
congestion, and enhances urban mobility by integrating
seamlessly into smart city infrastructure.
Motivation
Urban Parking Challenges
A solution is needed that can detect parking spaces automatically and help
drivers locate free spots easily.

Automation and Efficiency


The smart parking system brings efficiency, reliability, and ease of use.

Environmental Impact
Reducing the time spent searching for parking helps in saving fuel and
reducing carbon emissions.
Existing Model
Traditional Parking Systems:
Most current systems rely on manual operations, ticket-based entry, and often lack real-
time tracking of free parking spots.

Challenges include:
•No real-time data: Drivers have no way of knowing which spots are available.
•Inefficient space use: Parking spaces are often underutilized due to poor management and lack of
automation.
•Human dependence: Manual supervision is required, leading to errors and delays.
Proposed Model

•System Overview:

•Objective: Develop a parking space detection system using deep learning to address parking space management.

•Components:

•1. Detection of Parking Spots

•2. Detection of Cars

•3. Calculation of Intersection over Union (IoU)

•Detection of Parking Spots:

•- Approach 1: Identify parking spots using edge detection algorithms, but this may not work for all parking locations.

•- Approach 2: Use manual methods to mark parking regions in videos or images using Python's matplotlib library.
Mask R-CNN Algorithm
•Mask R-CNN extends Faster R-CNN by adding pixel-wise segmentation capability.
•It identifies object boundaries, making it suitable for detecting vehicles precisely within
parking spaces.
•Region Proposal Network (RPN):
Identifies potential object locations (vehicles).
•Bounding Box & Classification:
Detects cars in the image and places bounding boxes around them.
•Segmentation Mask:
Separates vehicles from the background for pixel-level accuracy.
• IoU (Intersection over Union):
Calculates the overlap between detected cars and parking spots to determine if
a spot is occupied.
Workflow of the System

Identify
Apply Mask R-
Input Capture Preprocessing Parking Output Display
CNN
Regions
The system
Predefined
overlays results
Detect vehicles parking spaces Users see live
Video feed or Frame extraction on the video
and mark their are checked for parking status via
static images of and video feed, showing
positions in real overlap with a dashboard or
the parking lot. normalization. whether a
time. detected application.
parking spot is
vehicles.
free or occupied.
Conclusion

•The proposed parking space detection system brings automation, scalability,


and efficiency to urban parking management.
•It integrates advanced object detection algorithms, image processing, and
real-time data visualization to enhance urban mobility.
•Potential future enhancements include deeper integration with mobile
applications for real-time parking spot reservation.
References
1. Saini M. Parking Space Detection Using Deep Learning. The Research
Nest.
https://medium.com/the-research-nest/parking-space-detection-using-deep-learning-
9fc99a63875e
2. Shaikh Z, Saini M, Shastri A. Automated parking space detection using
CNN.
https://www.researchgate.net/publication/322585370_Automated_parking_space_de
tection_using_CNN
3. Jia J, Chen L, Xu Y. Research Review on Parking Space Detection
Method.
https://www.researchgate.net/publication/348454518_Research_Review_on_Parking
_Space_Detection_Method

You might also like