Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection

INTRODUCTION: A robust method is proposed in this paper to detect helmet usage in two-wheeler riders to enhance road safety. OBJECTIVES: This involves a custom made dataset that contains 1000 images captured under diverse real-world scenarios, including variations in helmet size, colour, and light...

Full description

Saved in:
Bibliographic Details
Main Authors: Amrutha Annadurai, Manas Ranjan Prusty, Trilok Nath Pandey, Subhra Rani Patra
Format: Article
Language:English
Published: European Alliance for Innovation (EAI) 2025-03-01
Series:EAI Endorsed Transactions on Industrial Networks and Intelligent Systems
Subjects:
Online Access:https://publications.eai.eu/index.php/inis/article/view/7612
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850032479452266496
author Amrutha Annadurai
Manas Ranjan Prusty
Trilok Nath Pandey
Subhra Rani Patra
author_facet Amrutha Annadurai
Manas Ranjan Prusty
Trilok Nath Pandey
Subhra Rani Patra
author_sort Amrutha Annadurai
collection DOAJ
description INTRODUCTION: A robust method is proposed in this paper to detect helmet usage in two-wheeler riders to enhance road safety. OBJECTIVES: This involves a custom made dataset that contains 1000 images captured under diverse real-world scenarios, including variations in helmet size, colour, and lighting conditions. This dataset has two classes namely with helmet and without helmet. METHODS: The proposed helmet classification approach utilizes the Multi-Scale Deep Convolutional Neural Network (CNN) framework cascaded with Long Short-Term Memory (LSTM) network. Initially the Multi-Scale Deep CNN extracts modes by applying Single-level Discrete 2D Wavelet Transform (dwt2) to decompose the original images. In particular, four different modes are used for segmenting a single image namely approximation, horizontal detail, vertical detail and diagonal detail. After feeding the segmented images into a Multi-Scale Deep CNN model, it is cascaded with an LSTM network. RESULTS: The proposed model achieved accuracies of 99.20% and 95.99% using both 5-Fold Cross-Validation (CV) and Hold-out CV methods, respectively. CONCLUSION: This result was better than the CNN-LSTM, dwt2-LSTM and a tailor made CNN model.
format Article
id doaj-art-6fe4f8d3ca08476597fffafa3fdbddd5
institution DOAJ
issn 2410-0218
language English
publishDate 2025-03-01
publisher European Alliance for Innovation (EAI)
record_format Article
series EAI Endorsed Transactions on Industrial Networks and Intelligent Systems
spelling doaj-art-6fe4f8d3ca08476597fffafa3fdbddd52025-08-20T02:58:37ZengEuropean Alliance for Innovation (EAI)EAI Endorsed Transactions on Industrial Networks and Intelligent Systems2410-02182025-03-0112210.4108/eetinis.v12i2.7612Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet DetectionAmrutha Annadurai0Manas Ranjan Prusty1Trilok Nath Pandey2Subhra Rani Patra3Vellore Institute of Technology University Vellore Institute of Technology UniversityVellore Institute of Technology University The University of Texas at Arlington INTRODUCTION: A robust method is proposed in this paper to detect helmet usage in two-wheeler riders to enhance road safety. OBJECTIVES: This involves a custom made dataset that contains 1000 images captured under diverse real-world scenarios, including variations in helmet size, colour, and lighting conditions. This dataset has two classes namely with helmet and without helmet. METHODS: The proposed helmet classification approach utilizes the Multi-Scale Deep Convolutional Neural Network (CNN) framework cascaded with Long Short-Term Memory (LSTM) network. Initially the Multi-Scale Deep CNN extracts modes by applying Single-level Discrete 2D Wavelet Transform (dwt2) to decompose the original images. In particular, four different modes are used for segmenting a single image namely approximation, horizontal detail, vertical detail and diagonal detail. After feeding the segmented images into a Multi-Scale Deep CNN model, it is cascaded with an LSTM network. RESULTS: The proposed model achieved accuracies of 99.20% and 95.99% using both 5-Fold Cross-Validation (CV) and Hold-out CV methods, respectively. CONCLUSION: This result was better than the CNN-LSTM, dwt2-LSTM and a tailor made CNN model. https://publications.eai.eu/index.php/inis/article/view/7612Multi Scale CNNLong Short-Term MemoryDiscrete Wavelet TransformHelmet Detection
spellingShingle Amrutha Annadurai
Manas Ranjan Prusty
Trilok Nath Pandey
Subhra Rani Patra
Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection
EAI Endorsed Transactions on Industrial Networks and Intelligent Systems
Multi Scale CNN
Long Short-Term Memory
Discrete Wavelet Transform
Helmet Detection
title Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection
title_full Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection
title_fullStr Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection
title_full_unstemmed Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection
title_short Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection
title_sort single level discrete two dimensional wavelet transform based multiscale deep learning framework for two wheeler helmet detection
topic Multi Scale CNN
Long Short-Term Memory
Discrete Wavelet Transform
Helmet Detection
url https://publications.eai.eu/index.php/inis/article/view/7612
work_keys_str_mv AT amruthaannadurai singleleveldiscretetwodimensionalwavelettransformbasedmultiscaledeeplearningframeworkfortwowheelerhelmetdetection
AT manasranjanprusty singleleveldiscretetwodimensionalwavelettransformbasedmultiscaledeeplearningframeworkfortwowheelerhelmetdetection
AT triloknathpandey singleleveldiscretetwodimensionalwavelettransformbasedmultiscaledeeplearningframeworkfortwowheelerhelmetdetection
AT subhraranipatra singleleveldiscretetwodimensionalwavelettransformbasedmultiscaledeeplearningframeworkfortwowheelerhelmetdetection