0% found this document useful (0 votes)
30 views3 pages

Project Guidance

1. The project involves detecting helmets using YOLO detection and extracting license plate text using OCR from surveillance video of motorcycles. 2. Some challenges include the YOLO model having poor helmet detection accuracy compared to CNN, and detecting helmets when there are multiple riders on one motorcycle. 3. Potential applications of the ALPR system include highway toll collection, red light enforcement, border checkpoints, and parking management.

Uploaded by

xegago6945
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
30 views3 pages

Project Guidance

1. The project involves detecting helmets using YOLO detection and extracting license plate text using OCR from surveillance video of motorcycles. 2. Some challenges include the YOLO model having poor helmet detection accuracy compared to CNN, and detecting helmets when there are multiple riders on one motorcycle. 3. Potential applications of the ALPR system include highway toll collection, red light enforcement, border checkpoints, and parking management.

Uploaded by

xegago6945
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

What are the issues you have faced in this project ?

1.Detecting helmet by Yolo detection

2.extracting no plate text using OCR.

Technology stack:

YOLO algorithm,models ,V2,v3

OCR for recognising the text.

ML : is the study of computer algorithms, that improve automatically through experience & by the use
of data. It is seen part of AI.

open-CV : is a library of python package bindings designed to solve computer vision problems.

All the open CV arrray structures are converted to & from Numpy arrays makes it easier to integrate
other libraries that Use Numpy such as Scikit learn & matplodlib.

Drawbacks:

1. in Yolo model, helmet detection accuracy is very fast as compared to CNN model is poor with helmet
color.

2. in multiple Riders on a single motorcyclist it is difficult to detect helmet

3.tasks such as segmentation,classification and tracking can become even more difficult.

4. CCTV camera captures low resolution video some condition such as low light bad weather
complicating it.

Limitation

Limitations Due to time constraints Database could not be created.Record saved in Excel.

ALGORITHM

The algorithm we will use the YOLO algorithm.


YOLO = algorithm divides any given input image into SXS grid System. Each grid on the i/p image is
responsible for detection on object.

YOLO V1: 24 Convolutional layers that act as a feature extractor. 2 fully connected layers that are
responsible for classification of objeds & regression of bounding boxes.

YOLO V2: -remores all fully connected layers & uses anchor boxes to One pooling layer is removed to
predict bounding boxes. One pooling layer is removed to increase the resolution of o/p.

YOLO V3 = is a real time object detection algorithm that identifies specific objects in videos, live feeds
for images

o/p = for each spatial cell there are 125 versions.

Advantages :

1.smartphone can be used as input rather than using the webcam.

2.it can be open up a lot of possibilities as a smart phone can be carried out which can cover different
angles in it.

3.footage from mobile been up close can provide a clear image and most readable number plate for the
ocr which gives an accurate number.

4.it can be implemented in real time scenario using a webcam or CCTV as input.

Proposed System:

in this approach we will first detecting motorbikes Rider from the surveillance video by using
background subtraction as well as to object segmentation.

Conclusion/
Overview Report: from this, the system is developed for detecting the motorcyclist who are violating
the rules and regulation. system consists of three parts == detection of motorcycle , detection of helmet
and recognition licence plate.

Tensorflow: tensorflow is a free and open software source for data flow and programs that are
sequential across a range of task. it is used in ML application.

numpy: a standard list processing package. it provides the most efficient tool for the same members and
tools for working with these members.

Keras: is a open source software library that provides Python interface for neural input network.

Pandas: is a open software source library that provides a high performance data manipulation.

Applications

ALPR (automatic license plate recognition) system has many application which are Highway Tool
Collection, Red Light Violation enforcement, Border & Custom Checkpoints, Management of parking
areas and to display the available space in parking.

Future scope:

In future, Image capturing system will install. Camera will place on door, when car will arrive camera
capture picture of front of car then localize the number plate and do further recognition process. If
number plate is authorized then door will open otherwise an alarm will ring.

You might also like