0% found this document useful (0 votes)
67 views6 pages

Deep Learning Model For Accident Detection in Smart Cities Using Computer Vision

The document describes a deep learning model for detecting accidents in smart cities using computer vision techniques. The model uses a convolutional neural network (CNN) that is trained on a large dataset of accident images and videos to recognize features that indicate an accident. Once trained, the model can analyze live video feeds from traffic cameras in real-time to detect accidents and automatically alert emergency services, providing them with timely information to reduce response times. Experimental results showed the model achieved around 89% accuracy in validating its ability to detect accidents from camera footage. The system aims to help emergency services respond more quickly to save lives and reduce injuries from traffic accidents.

Uploaded by

amine
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
67 views6 pages

Deep Learning Model For Accident Detection in Smart Cities Using Computer Vision

The document describes a deep learning model for detecting accidents in smart cities using computer vision techniques. The model uses a convolutional neural network (CNN) that is trained on a large dataset of accident images and videos to recognize features that indicate an accident. Once trained, the model can analyze live video feeds from traffic cameras in real-time to detect accidents and automatically alert emergency services, providing them with timely information to reduce response times. Experimental results showed the model achieved around 89% accuracy in validating its ability to detect accidents from camera footage. The system aims to help emergency services respond more quickly to save lives and reduce injuries from traffic accidents.

Uploaded by

amine
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Journal of Harbin Engineering University

ISSN: 1006-7043 Vol 44 No. 6


June 2023

Deep Learning Model for Accident Detection in Smart Cities using Computer
Vision

Jidgnesh Sathishchandra Wani, Keerthi A., Kuriti Preethi Naidu, Shanthi M. B.


Department of CSE, CMR Institute of Technology, Bengaluru
[email protected], [email protected], [email protected], [email protected]

Abstract: In recent years, the concept of smart cities has gained increasing attention, driven by the need to
make urban environments safer, more sustainable, and more efficient. One of the key challenges in achieving
this vision is the management of traffic, which is a major cause of accidents and delays in urban areas.
Traditional methods of accident detection rely on manual reporting or sensor-based systems, which are often
slow and imprecise. To address this challenge, researchers have proposed using computer vision techniques to
automatically detect accidents in real-time. We have developed a deep learning model; CNN is capable of
analyzing real-time video feeds from surveillance cameras installed in strategic locations throughout a city. In
this paper, we will describe the proposed model in detail, including the architecture of the CNNs, the training
process, and the real-time accident detection algorithm. Experimental observations have shown around 89%
of validation accuracy in detecting the accidents based on the live video feed from cameras.
Keywords: smart cities, computer vision, deep-learning, sensor-based, video-feeds

I. Introduction location of the accident, leading to further delays


The concept of smart cities has given rise to the in reaching the scene.Therefore, there is a need
development of smart solutions for real time for a system that can automate the process of
problems. Traditional methods of accident accident detection and response. The system
detection rely on manual reporting or sensor- should be capable of detecting accidents in real-
based systems, which are often slow and time, accurately identifying the location and
imprecise. To address this challenge, researchers severity of the accident, and automatically
have proposed using computer vision techniques notifying the relevant authorities. This would
to automatically detect accidents in real-time. This ensure that emergency services are alerted
approach has the potential to save lives and immediately, reducing the response time and
reduce injuries in accidents, as it provides providing prompt assistance to those in need.
emergency services with real-time information
about the location and severity of the incident. II. Related Study
Every year, numerous road accidents The research study shows the use of Artificial
occur all around the world, resulting in the loss of Intelligence in detection of accidents for quick
countless lives and serious injuries. Delayed response in smart cities. Here is an overview of
response from emergency services often results in recent research on DL models for accident
additional loss of life and property. Therefore, detection in Smart Cities. Y. Chen et al. [1],
there is a pressing need for a system that can proposed a DL model for real-time accident
detect accidents in real-time and provide detection in urban traffic scenes. The model
emergency services promptly.The current system employs a Faster R-CNN algorithm and ResNet-50
for accident detection and emergency response is as a backbone network to detect and classify
largely dependent on human intervention. It relies accident images into three categories: collision,
on individuals or passersby to alert authorities of fire, and smoke. Hang Shi et al. [2] have developed
an accident, which often leads to a delay in a YOLOv4 method for object detection and object
response time. In some cases, the severity of an tracking based on Kalman filter coupled with the
accident can goes unnoticed entirely, and no Hungarian algorithm for association, and accident
assistance is provided. Furthermore, emergency detection by trajectory conflict analysis. They have
services are often unable to locate the exact got an accuracy of 93.1%.Computer Vision Based

222
Journal of Harbin Engineering University
ISSN: 1006-7043 Vol 44 No. 6
June 2023

Approach to Detection and Classification of Traffic III. System Design and Implementation
Incidents was employed by RazanSendi et al. [3]. The proposed system aims to predict the accident
The vehicle detection and tracking model assigned and alert the nearest emergency services to
a distinct identification number (ID) to each provide timely first aid to the victim and
vehicle and combined the YOLOv5 object detector potentially save their life. To achieve this goal,
with the DeepSORT tracker to detect and monitor Deep Learning techniques based on Convolutional
the movements of the vehicles. The mean average Neural Networks (CNN) will be used. The
precision (mAP) of this model was 99.2%. A human implementation of this system involves collecting a
operator or emergency services would be notified large dataset of accident images or videos, pre-
in the event of an accident utilising a computer processing the data, training the CNN model, and
vision-based deep learning model for accident deploying it to detect accidents and alert
detection in smart cities. emergency services. Fig. 1 shows the system
architecture for the proposed system.

Fig. 1. System architecture for the proposed system.

The system architecture for accident detection recognize the features that are indicative of an
using Convolutional Neural Networks (CNN) accident, such as damaged vehicles or debris on
typically involves the following components. the road.
A.Data Collection: A large dataset of images D. Model Training: The CNN model needs to be
and/or videos of accident scenes needs to be trained using the prepared dataset. This involves
collected. This dataset should include a variety of optimizing the model's weights and biases to learn
accidents, including different types of vehicles, to recognize the features that are indicative of an
different road conditions, and different degrees of accident.
damage. E. Model Evaluation: Once the model has been
B. Data Pre-processing: The collected data needs trained, it needs to be evaluated using a separate
to be pre-processed to ensure that it is suitable for set of images and/or videos that were not used
use with the CNN. This may involve tasks such as during training. This is to evaluate the model's
resizing images, normalizing pixel values, and performance on unseen data and ensure that it is
separating the dataset into training, validation, able to generalize to new accident scenarios.
and testing sets. F. Deployment: Once the model has been tested
C. CNN Model Architecture: The CNN model and its performance is satisfactory, it will be
architecture needs to be designed, which typically deployed in a real-world scenario. It would be
involves choosing the number and type of layers, integrated into a traffic monitoring system that
the size of the filters, and the activation functions. analyses live video feeds from traffic cameras to
The model should be designed to learn to detect accidents and alert emergency services.

223
Journal of Harbin Engineering University
ISSN: 1006-7043 Vol 44 No. 6
June 2023

Fig. 2 Shows the Workflow Of The Proposed System.

The system receives input from cameras installed accident.The system alerts the nearest police and
in the area that monitor traffic and detect hospital to provide timely first aid and potentially
accidents. It uses computer vision techniques, such save the victim's life.The system tracks the victim's
as object detection and image segmentation, to location and provides real-time updates to
identify accidents in the input data.Once an emergency services until they arrive on the scene.
accident is detected, the system uses a CNN model The graphs given in Fig.3 and Fig.4 shows the
trained on accident images to predict the type of training loss verses training accuracy and
accident, such as car crash or any other type of validation loss verses validation accuracy.

Fig.3 Training loss V/S Training accuracy Fig.4 Validation Loss V/S Validation Accuracy

IV. Results And Discussion the size and position of its wheels, the presence of
When a photo is analyzed using a Convolutional headlights or taillights, and other details that help
Neural Network (CNN) algorithm for car detection, distinguish cars from other objects in the image. It
the algorithm will look for features in the image is shown in Fig. 5 in which car is detected and Fig.
that are commonly associated with cars[6]. These 6 in which accident is not detected.
features might include the shape of the car's body,

Fig. 5 Car is detected. Fig. 6 Accident is not detected

224
Journal of Harbin Engineering University
ISSN: 1006-7043 Vol 44 No. 6
June 2023

A video is divided into individual frames to identify non-accident scenarios, so that it can learn to
if an accident has occurred.When a video is distinguish between the two. The algorithm may
analyzed using a Convolutional Neural Network also need to be programmed to recognize
(CNN) algorithm for accident detection, the different types of accidents, such as vehicle
algorithm will look for patterns in the video frames collisions, pedestrian accidents, or fires.
that indicate the presence of an accident. These Once the CNN algorithm has detected an accident
patterns might include sudden changes in the in the photo, it can be used to trigger an alert or
motion of objects in the video, the appearance of notification to a human operator or emergency
smoke or debris, or other visual cues that suggest services. The alert can include details such as the
a collision or other type of incident has location of the accident, the number of vehicles
occurred[14]. To detect accidents, the CNN involved, and any other relevant information that
algorithm will likely need to be trained on a large can assist in responding to the incident. Fig. 7, 8, 9
dataset of videos that include both accident and shows the different frames used.

Fig. 7 Frame 1

Fig. 8 Frame 2 Fig. 9 Frame 3


An alert message sent to emergency medical immediate medical attention[11]. The message is
services during an accident is a communication usually initiated by a witness to the accident or
that is sent to notify medical personnel of a someone who has come across the scene of the
serious incident or accident that requires accident.The alert message may include details

225
Journal of Harbin Engineering University
ISSN: 1006-7043 Vol 44 No. 6
June 2023

such as the location of the accident, the nature References:


and severity of injuries sustained by the individuals [1] Computer Vision based Traffic Monitoring and
involved, and any other relevant information that Analyzing from On Road Videos By T.M. Amir-Ul-
could assist medical personnel in providing timely Haque Bhuiyan, Mrinmoy Das & Md. Shamim Reza
and appropriate care. The message is usually sent Sajib Bangladesh University of Business &
to a centralized emergency medical services (EMS) Technology (BUBT).
dispatch center, which is responsible for [2] Automatic Traffic Surveillance using Video
coordinating the response to the accident.Upon Tracking Bhushan Nemade Department of
receipt of the alert message, the EMS dispatch Information Technology, Thakur College of
center will typically mobilize appropriate medical Engineering and Technology. Mumbai 400101.
resources, such as ambulances, paramedics, and [3] Accident Detection using Deep Learning: A
other first responders, to the location of the Brief Survey Renu 1 , Durgesh Kumar Yadav 2*
accident. The medical personnel will then assess ,Iftisham Anjum 3 and Ankita 4 1 Assistant
the situation, provide any necessary medical Professor, Department of Computer Science and
treatment, and transport the injured individuals to Engineering, Greater Noida, Uttar Pradesh, India.
the nearest hospital or medical facility. [4] Vehicle Detection and Tracking for Traffic
Surveillance Applications: A Review Paper Anshul
III.Conclusion And Future Scope Vishwakarma , Amit Khare.
In conclusion, the proposed system for reducing [5] Accident Detection and Emergency Alerting
fatality in road traffic accidents utilizes CNN-based System for Road Safety Dilipkumar
systems to analyze video footage in real-time and A. Borikar1* and Manoj B. Chandak2 1,2Shri
detect potential accidents. By automatically Ramdeobaba College of Engineering and
alerting emergency services, this system aims to Management, Nagpur, Maharashtra, India.
facilitate a more rapid response and potentially [6] Car crash detection using ensemble deep
save lives. The use of CNNs offers advantages such learning and multimodal data from dashboard
as improved accuracy and speed in accident cameras(Jae GyeongChoia1Chan
detection compared to traditional methods. WooKonga2GyeonghoKima3 SunghoonLim)
However, it is important to acknowledge that Science Direct.
there are still limitations and challenges associated Computer Vision-based Accident Detection in
with this approach. One key challenge is the Traffic Surveillance IEEE.
requirement for large datasets to effectively train [7] Gowshika B, MadhuMitha, G,Jayashree,Vehicle
the CNN models. Sufficient and diverse data is accident detection system usingGPS, GSM modem,
necessary to ensure that the system can accurately IRJET, 2019.
identify various types of accidents and distinguish [8] Sayanee Nanda, Harshada Joshi, Smitha
them from false positives. Seamless Khairnar, An IOT Based Smart System for
communication and coordination between the Accident Prevention and Detection,IEEE,2018.
accident detection system and emergency service [9] Norsuzila Yaacob, AinnurEizaAzhar, Azita
providers need to be established to ensure timely LailyYusofl, Suzi SerojaSarnin,Darmawaty Mohd Ali
response and appropriate allocation of resources. and Ammar Anuar, Real Time Wireless Accident
With further advancements in technology and Tracker using Mobile Phone,IEEE,2017.
improvements in data availability, these systems [10] Video-Based Traffic Data Collection System for
have the potential to enhance road safety and Multiple Vehicle Types ShuGuang
minimize the fatality rate associated with road Li, Yu Hongkai, Zhang Jingru.
traffic accidents. Continued research and [11] Almeida, T. A., Abreu, S., Silva, D. F., & Nunes,
development in this field are essential to U. (2019). A survey on accident detection and
overcome the existing limitations and maximize management using wireless sensor networks and
the effectiveness of such systems in real-world machine learning techniques. Sensors, 19(19),
scenarios. 4183.

226
Journal of Harbin Engineering University
ISSN: 1006-7043 Vol 44 No. 6
June 2023

[12] Chou, J. S., Lai, T. H., & Huang, C. C. (2020).


Review on real-time accident detection and
warning systems for intelligent transportation
systems. Applied Sciences, 10(6), 2087.
[13] Zhang, G., Sun, W., Wang, H., & Sun, H.
(2020). An intelligent emergency accident
detection algorithm based on a deep learning
model. Sensors, 20(3), 832.
[14] Noh, H., Kim, D., Kim, D., & Park, J. (2019).
Accident detection using deep learning-based
video analysis for intelligent transportation
systems. Electronics, 8(6), 674.
[15] Xu, Y., Mao, Y., Zhang, L., Tang, M., & Wang, J.
(2021). Review on accident detection and
intelligent rescue systems based on computer
vision. Sensors, 21(4), 1223.

227

You might also like