No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Helmet Detection using Machine Learning Approach
... Triple seat spot detection, leveraging computer vision and machine learning, assists in enforcing traffic regulations, improving motorcycle safety. [1]The reluctance among motorcyclists to wear helmets has heightened head and brain injuries in accidents. This [1] study introduces a robust framework for detecting helmet-less riders, utilizing YOLOv3 for initial rider detection and a novel CNN architecture for helmet detection. ...
... [1]The reluctance among motorcyclists to wear helmets has heightened head and brain injuries in accidents. This [1] study introduces a robust framework for detecting helmet-less riders, utilizing YOLOv3 for initial rider detection and a novel CNN architecture for helmet detection. When compared to other CNN-based techniques, the model's efficacy is assessed using a variety of traffic videos, demonstrating encouraging outcomes. ...
The escalating motorcycle accident rates highlight the pressing need for improved safety measures. Helmets, a crucial safety gear, are often neglected, contributing significantly to fatalities. This paper addresses the pervasive issue of noncompliance with motorcycle safety rules, focusing on helmet usage and triple riding. Existing systems for monitoring lack precision, prompting our proposed Bike Traffic Violation System. Leveraging Haar Cascade and YOLOv3 models, it identifies motorcycles, detects riders without helmets, instances of triple riding, and even empty parking spots with unprecedented accuracy. The Machine Learning component employs a Support Vector Classification model, bolstered by 4-fold cross-validation, ensuring robustness. This innovative system provides real-time insights into traffic violations, enabling prompt interventions by law enforcement agencies. It overcomes manual identification shortcomings, offering a comprehensive solution for enforcement and awareness campaigns. In summary, our Bike Traffic Violation System not only advances automated traffic rule monitoring but introduces a novel methodology for precise detection, significantly contributing to enhanced road safety and accident prevention.
... Hu et al. [9] developed a model to classify and detect vehicles in video footage using Histogram of Oriented Gradients (HOG) method. However, their study focused solely on identifying the vehicles and did not address the detection or analysis of the people traveling inside those vehicles.Vaishali et al. [10] proposed machine learning based helmet detection model. The system uses sensors and a machine learning-based cascade classifier with HAAR features to check if a rider is wearing a helmet. ...
To restraints the rate of fatality due to accident, a powerful and adequate implementation of traffic rules and continuous monitoring is required. Traffic Rule Violation Detection (TRVD) system aims to identify the traffic rule violation – triple riding and ensures that the rule must be followed 24*7 without any human intervention. To detect the traffic rule violation, deep learning based single shot detection algorithm is utilized. YOLO (You Only Look Once) algorithm used for detection of two wheeler and number of persons riding on a motorcycle, the system detect and classify a person is following a traffic rule strictly or not. The suggested method trains a model on a datasets that combines custom images with publicly available datasets. This approach is very effective at accurately detecting traffic rule violations related to triple riding, whether it's a single rider or multiple riders on the bike.Furthermore, to address the issue of class imbalance, data augmentation techniques were utilizedto increase the variation in training data. This strengthen the model's effectiveness in applying to practical scenarios. Different YOLO family algorithm has been utilized for development of detection model. The YOLOv8 model was tested on a total of 80 images and detection accuracy exhibited an F1 score, precision and mAP@50 of 76.4%, 73.5% and 81.6% respectively for all classes. We manually tested triple riding traffic rule violation using our proposed algorithm and found that the system gives 92.7% of accuracy. These findings highlight the potential of proposed model, thus fostering safer motorcycling practices.
... 1. Helmet Detection Model and Video Loading: Loads a YOLO model for helmet detection and opens a video file [14] . Detection: Identifies helmets in each frame, drawing green bounding boxes labeled "Helmet" with confidence scores. ...
... If the bike rider is in emergency condition then by using GSM and GPS sharing of the exact location is automatically existed. The other set of sensors is fixed to the helmet whether the user is drunken or not for their security purpose [5]. Along with other types of photoelectric cells speed limit sensors are fixed to the helmet module for decreasing the speed of motor vehicles. ...
Welcome to "Connected Horizons: Exploring IoT Applications in Infrastructure, Agriculture, Environment, and Design," a groundbreaking exploration edited by Professor Jagdish H. Godihal, Professor in Civil Engineering at Presidency University, Bengaluru. In the ever-evolving landscape of technology, the Internet of Things (IoT) has emerged as a transformative force, reshaping industries and revolutionizing the way we interact with our environment. This anthology brings together a diverse array of experts, each contributing their unique insights into the potential of IoT across various domains. From the deployment of IoT in animal healthcare and location tracking to its vital role in modern power systems, the chapters in this volume offer a comprehensive look at the breadth and depth of IoT applications. Whether it's the implementation of telemedicine applications for pandemics or the optimization of wastewater treatment plants using IoT-based solutions, the possibilities are as vast as they are innovative. As we analyse deeper into the book, we encounter discussions on IoT in logistics, agriculture, smart buildings, and beyond. The intersection of IoT with artificial intelligence opens up new avenues for efficiency and automation, while practical applications in smart homes and cities showcase the tangible benefits of connected technologies. Through case studies and practical examples, this book not only illuminates the current state of IoT but also offers glimpses into its future potential. From intelligent street lighting to safe riding solutions and battery optimization for electric vehicles, the chapters in this anthology exemplify the transformative power of IoT in shaping a smarter, more sustainable world. As we start this exploration journey, we welcome readers to join us in uncovering the boundless opportunities presented by IoT. Together, let us navigate the connected horizons and chart a course towards a future defined by innovation, collaboration, and technological advancement. I am delighted to express my sincere gratitude to all the authors for their invaluable contributions to this book on Connected Horizons: Exploring IoT Applications in Infrastructure, Agriculture, Environment, and Design. Special appreciation is extended to Iterative International Publishers (IIP) and Selfypage Developers Pvt Ltd., located in Chikkamagaluru, Karnataka, India – 577102, for their dedicated efforts in bringing this volume to fruition. I would like to acknowledge the remarkable contributions of several colleagues at Presidency University, including Dr. Nakul Ramanna, Professor and Head of the Department of Civil Engineering; Dr. Ashok Itagi, Head of the School of Design; Dr. Manohar Joshi, Professor and Head of the Department of Electrical and Electronics Engineering, Dr. Mahaboob Pasha, Professor and Head of the Department of Physics, Dr. Shashikala, A R Professor and Head of the Department of Chemistry, Dr. Gopal Shyam, Professor CSE, Presidency University, Bengaluru; Dr. Ganesh Hegde, Professor Civil Engineering, Goa College of Engineering, Farmagudi, Goa, Mrs Himmi Gupta, Assistant Professor, Civil Engineering Department, NITTTR, Sector 26, Chandigarh. Dr. Rashmita Srinivasan, Associate Dean (Academics), Civil Engineering Department, Maharashtra Institute of Technology (Autonomous), Aurangabad. Dr. Sreenivasappa, Professor of Electronics Engineering; Dr. Syed Abid Hussain, Associate Professor, School of Commerce, Galiveeti Poornima, Assistant Professor, Department of Computer Science and Engineering, School of Engineering; Dr. Ravi Angadi, Assistant Professor, EEE Department, Madhusudana M, Assistant Professor, School of Design [Fashion Design],Presidency University, Bengaluru, Dr. A Jency Priyadharshany, Asst. Professor, School of Commerce, Presidency University, Bengaluru and Dr. Madhavi, T, Dr. Venkatesh Raju K, Associate Professor, Civil Engineering and Dr. Divya Rani, Associate Professor, Electronics and Communication Engineering. Dr. Mohammed Mujeer Ulla, Assistant Professor- Selection Grade, Presidency University, Preethi, Assistant Professor- Senior Scale and Sapna R, Assistant Professor, Manipal Institute of Technology, Bengaluru. Mr. Santhosh M B, Dr Shwetha, Mr. Bhavankumar, Mr. Gopalkrishnan, Mrs. Divya Nair, Mrs. Sowmyashree, Mr. Dayalan, Mr. Ajay, Mr. Karthik Assistant Professors, Department of Civil Engineering, Presidency University, Bengaluru; Mr. Harshavardhana, Mr. Jithendra, Mr.Vinayak Babu and Mr. Bibang, Mrs. Chetana, Mrs. Manjula, Mrs. Padmavathi and Mrs. Chaithra, Mrs. Shobha, Research Scholars, contribution and support have been instrumental in shaping this publication. Gratitude is extended to the management authorities and the key officials of Presidency University, Bengaluru, for their continued support and encouragement throughout this endeavour. Special thanks to my wife, Kavita, and daughter, Saatvika, for their invaluable suggestions, unwavering support, insightful feedback, and constant motivation throughout the journey of completing this book.
... Their hierarchical architecture allows for progressive refinement of object localization and classification, ultimately leading to more accurate helmet detection results. [31,32]. ...
In India, helmets symbolize safety and civic responsibility, bearing cultural significance. However, a 22% increase in accidents and a 17.5% rise in fatalities in 2022-23 underscore the critical importance of helmet compliance beyond legal mandates. Non-compliance not only elevates the risk of injuries and fatalities but also entails legal consequences. Notably, 47,000 Indians died in 2021 due to not wearing helmets, emphasizing the pivotal role of helmet usage in road safety. This research focuses on improving motorcycle helmet detection to ensure compliance and reduce the risk of fatal head injuries for riders, extending its impact beyond geographical limits. While our dataset predominantly draws from Sivasagar, a district in Assam, India, the scope of our research is universally applicable. We employed a comprehensive methodology, comprising data collection, preprocessing, and YOLOv5 model training using the Darknet framework, testing, and evaluation. Analysis of the original YOLOv5 algorithm's performance using Precision-Recall (PR) curves resulted in mAP values of 85.9% for helmets, 88.1% for human heads, and an average of 87%. Subsequently, the proposed YOLOv5 algorithm, achieving mAP values of 93% for helmets, 96.8% for human heads, and a remarkable 94.9% average mAP, demonstrated significant improvements. Comparison revealed a consistent 7–8.5% higher mAP for helmet and human head detection, underscoring the efficacy of the proposed approach in improving detection capabilities. This research contributes to the broader field of computer vision and its practical applications, particularly in enhancing road safety and averting head injuries among riders, irrespective of their location.
... The key disadvantages are that consumption time is long and it necessitates a greater focus on global information. In 2019, Liao et al. [21] published M-LAP, a unique approach for combining characteristics from various scales to improve the efficacy of glaucoma analysis. Furthermore, the provided strategy produced glaucoma activation, bridging the gap between global semantical analysis and precise locations. ...
Diagnosis and diagnosis of glaucoma progression remain difficult. Artificial intelligence-based techniques have the potential to improve and standardize glaucoma examination, however because of the multimodal and changeable nature of the diagnosis, developing these algorithms is difficult. Most algorithms are now focused on a single imaging modality, namely screening and diagnosis using fundus pictures or optical coherence tomography images. In our review of the literature, we found no research that evaluated the use of artificial intelligence for treatment response prediction, and no studies that conducted prospective testing of their algorithms. Another barrier to the development of artificial intelligence-based solutions is a lack of data and agreement on diagnostic criteria. Although research on the use of artificial intelligence for glaucoma is promising, more research is required to develop clinically usable tools. Current glaucoma detection convolutional neural networks (CNNs) are all based on spatial data embedded in an image. We created a hybrid CNN and recurrent neural network (RNN) that extracts not only the spatial but also the temporal characteristics encoded in a fundus image. A total of 1810 fundus photos and 295 fundus images were used to train a CNN and a CNN-LSTM-RNN combination. In differentiating glaucoma from healthy eyes, the combined CNN/RNN model achieved an average F-measure of 94.3%. A detecting system is required to aid in the early identification of glaucoma. The researchers propose employing deep learning technology to detect and forecast glaucoma before symptoms arise in this study. The results are contrasted with deep learning-based convolution neural network classification techniques. The suggested model has an accuracy of 98.21% when used for training and an accuracy of 96.34% when used for testing. According to all evaluations, the newly proposed paradigm is more effective than the one that is already in use.
Traffic signs are vital for maintaining smooth traffic flow and preventing accidents, conveying crucial information to drivers through pictorial representations. However, drivers often overlook these signs, leading to potential accidents due to factors like difficulty in focusing, fatigue, and environmental conditions. To tackle this issue, innovative machine-learning techniques have been developed. While traditional traffic sign detection relied on conventional object detection methods, recent advancements, particularly the YOLOv5 (’You Only Look Once’) algorithm, have significantly improved accuracy and speed. YOLOv5, with its grid structure, is renowned for achieving a remarkable accuracy rate of 90.25%. Our proposed solution leverages YOLOv5 for object detection, dividing images into a grid to locate objects efficiently. The technology not only enhances accuracy but also provides real-time analysis and alerts users through audio signals. In future work, our goal is to further enhance the model’s accuracy, expand its capabilities for detecting various signs, and implement it in real-time applications.
Helmet detection ensures safety in industries like construction, manufacturing and bike driving. This technology uses advanced image processing and machine learning algorithms to identify whether an individual is wearing a helmet. By preventing head-related accidents, helmet detection promotes a culture of safety among employees and helps companies comply with safety regulations. This study, based on the You Only Look Once model (YOLO) for helmet detection, uses transfer learning to train the model on a dataset of helmet images and fine-tune it to detect helmets. This approach involves starting with a pre-trained YOLO model on a large general object detection dataset and then, fine-tuning the model on a smaller dataset of helmet images. It is highly accurate and effective in handling complex lighting conditions and perspectives, making it less sensitive to image orientation and lighting conditions compared to traditional methods. YOLO is also scalable and adaptable. Experimental results demonstrate the method is effective in helmet detection in terms of precision, recall and mAP metrics.
Articulating an individual behavior analysis from a video stream is considered by most researchers. It has its own applications in the field of computer vision-based Human activity recognition (HAR). HAR is the most widely utilized service in many of the systems across surveillance systems, healthcare, online education, and many more areas. Nowadays equally there is also increased interest in the researchers across the HAR community for works to be considered when multiple faces have to be recognized and predict the activity. This paper details the contribution made by the researchers in this aspect recently and also paper gives a comprehensive analysis of the methods adopted and concludes by examining the accuracy of the various work contributed. At last, this paper additionally is given the future directions to deal with this application.
The expanding capabilities of today’s microcontrollers and other devices lead to an increased utilization of these technologies in diverse fields. The automation and issue of remote control of moving objects belong to these fields. In this project, a microcontroller Raspberry Pi 2B was chosen for controlling DC motors and servo-motors. This paper provides basic insight into issue of controlling DC motors and servo-motors, connection between Raspberry and other components on breadboard and programming syntaxes for controlling motors in Python programming language.
Today image processing are used in various techniques, this paper presents the implementation of image processing operations on Raspberry Pi. The Raspberry Pi is a basic embedded system and being a low cost a singleboard
computer used to reduce the complexity of systems in real time applications. This platform is mainly based on
python. Raspberry pi consist of Camera slot Interface (CSI) to interface the raspberry pi camera. Here, the Dark and
Low contrast images captured by using the Raspberry Pi camera module are enhanced in order to identify the particular region of image. This concept is used in the real time application of MAV, The MAVs are basically used to capture images and videos through the Raspberry pi camera module. Because of its credit card sized (small) and less weight in the design. However, the image captured by MAVs will consist of unwanted things due to atmospheric conditions; hence it is necessary to remove noise present in the MAVs images.
This paper presents a framework for automatic detection of bike-riders without helmet using surveillance videos in real time. The proposed approach first detects bike riders from surveillance video using background subtraction and object segmentation. Then it determines whether bike-rider is using a helmet or not using visual features and binary classifier. Also, we present a consolidation approach for violation reporting which helps in improving reliability of the proposed approach. In order to evaluate our approach, we have provided a performance comparison of three various feature representations for classification. The experimental results show detection accuracy of 93.80% on the real world surveillance data. It has also been shown that proposed approach is computationally less expensive and performs in real-time with a processing time of 11.58 ms per frame.
Motorcycle accidents have been rapidly growing throughout the years in many countries. Due to various social and economic factors, this type of vehicle is becoming increasingly popular. The helmet is the main safety equipment of motorcyclists, however many drivers do not use it. The main goal of helmet is to protect the drivers head in case of accident. In case of accident, if the motorcyclist does not use can be fatal. This paper aims to propose a system for detection of motorcyclist without helmet. For this, we have applied the circular Hough transform and the Histogram of Oriented Gradients descriptor to extract the image attributes. Then, the MultiLayer Perceptron classifier was used and the obtained results were compared with others algorithms. Traffic images were captured by cameras from public roads and constitute a database of 255 images. Indeed, the algorithm step regarding the helmet detection accomplished an accuracy rate of 91.37%.
Although motorcycle safety helmets are known for preventing head injuries, in many countries, the use of motorcycle helmets is low due to the lack of police power to enforcing helmet laws. This paper presents a system which automatically detect motorcycle riders and determine that they are wearing safety helmets or not. The system extracts moving objects and classifies them as a motorcycle or other moving objects based on features extracted from their region properties using K-Nearest Neighbor (KNN) classifier. The heads of the riders on the recognized motorcycle are then counted and segmented based on projection profiling. The system classifies the head as wearing a helmet or not using KNN based on features derived from 4 sections of segmented head region. Experiment results show an average correct detection rate for near lane, far lane, and both lanes as 84%, 68%, and 74%, respectively.
This paper describes a machine learning approach for visual object detection which is capable of processing images extremely rapidly and achieving high detection rates. This work is distinguished by three key contributions. The first is the introduction of a new image representation called the "integral image" which allows the features used by our detector to be computed very quickly. The second is a learning algorithm, based on AdaBoost, which selects a small number of critical visual features from a larger set and yields extremely efficient classifiers. The third contribution is a method for combining increasingly more complex classifiers in a "cascade" which allows background regions of the image to be quickly discarded while spending more computation on promising object-like regions. The cascade can be viewed as an object specific focus-of-attention mechanism which unlike previous approaches provides statistical guarantees that discarded regions are unlikely to contain the object of interest. In the domain of face detection the system yields detection rates comparable to the best previous systems. Used in real-time applications, the detector runs at 15 frames per second without resorting to image differencing or skin color detection.
Motorcycle accidents have been rapidly growing throughout the years in many countries. Due to various social and economic factors, this type of vehicle is becoming increasingly popular. The helmet is the main safety equipment of motorcyclists, but many drivers do not use it. If an motorcyclist is without helmet an accident can be fatal. This paper aims to explain and illustrate an automatic method for motorcycles detection and classification on public roads and a system for automatic detection of motorcyclists without helmet. For this, a hybrid descriptor for features extraction is proposed based in Local Binary Pattern, Histograms of Oriented Gradients and the Hough Transform descriptors. Traffic images captured by cameras were used. The best result obtained from classification was an accuracy rate of 0.9767, and the best result obtained from helmet detection was an accuracy rate of 0.9423.
Helmet Detection using Machine Learning and Automatic License Plate Recognition
- Lokesh Allamki
- Manjunath Panchakshari
- Ashish Sateesha
- K S Pratheek
Machine vision techniques for motorcycle safety helmet detection
- R Waranusast
- N Bundon
- V Timtong
- C Tangnoi
- P Pattanathburt
Image processing based intruder Detection using Raspberry PI
- S Adhav
- R Manthira Moorthi
- G Prethija
Smart helmet-Intelligent Safety For Motorcyclist Using Raspberry Pi And OpenCv
- Bhagyashree Shabrin
- Jagadish Nikharge
- M Maithri
- T Poojary
- B Pooja
- Adana
Smart helmet-Intelligent Safety For Motorcyclist Using Raspberry Pi And OpenCv
- Shabrin
Helmet Detection using Machine Learning and Automatic License Plate Recognition
- Allamki
Image processing based intruder Detection using Raspberry PI
- Adhav