https://github.com/parth1899/auto-blackbox-3d

https://github.com/parth1899/auto-blackbox-3d

Last synced: 9 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/parth1899/auto-blackbox-3d
• Owner: parth1899
• Created: 2024-05-18T19:53:58.000Z (about 2 years ago)
• Default Branch: main
• Last Pushed: 2024-08-06T04:16:02.000Z (almost 2 years ago)
• Last Synced: 2024-08-06T06:51:06.589Z (almost 2 years ago)
• Language: JavaScript
• Homepage: https://auto-black-box-3-d.vercel.app
• Size: 1.47 MB
• Stars: 1
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Auto-BlackBox-3D

## Employing Black Box Mechanism for Investigation and Analysis of Road Accidents

## Overview

This project presents a novel system for accident analysis using advanced sensor technology. The system gathers real-time acceleration, gyroscope, and GPS data to create detailed 3D models of vehicle orientation during accidents. By employing machine learning techniques and data visualization tools, the system enhances post-accident analysis and provides valuable feedback for improving vehicle design and safety.

## Features

- **Data Collection**: Utilizes MPU6050 and GPS sensors to gather acceleration, gyroscope, and location data.

- **Anomaly Detection**: Employs autoencoders to detect anomalies in acceleration values.

- **3D Modeling**: Uses Three.js to create real-time 3D models of vehicle orientation.

- **Data Visualization**: Implements Plotly.js for comprehensive data visualization.

- **Location Tracking**: Integrates OpenStreetMap API for live GPS tracking.

## System Architecture

![System Architecture](images/system-architecture.png)

## Methodology

1. **Data Collection**: Sensors capture acceleration, gyroscope, and GPS data.

2. **Data Processing**: NodeMCU processes and transmits data to the cloud.

3. **3D Modeling**: Three.js visualizes vehicle orientation in 3D.

4. **Anomaly Detection**: Autoencoders identify anomalies in the data.

5. **Data Visualization**: Plotly.js generates graphs for data analysis.

## Results

- **3D Model Rendering**: Real-time visualization of vehicle orientation during accidents.

- **Anomaly Detection**: Identified 581 anomalies with a 94.99% accuracy rate.

- **Data Visualization**: Comprehensive graphs showing acceleration and gyroscope data.

### Homepage of Flask Application

![Homepage of Flask Application](images/homepage.png)

### Hardware Connections

![Hardware Connections](images/hardware.png)

### Viewing 3D Model

![Viewing 3D Model](images/view-3D-Model.png)

### Anomalies

![Anomalies](images/anomalies.png)

### PCA Projection of Sensor Readings

![PCA Projection of Sensor Readings](images/pca.png)

### Tracking Live Location

![Tracking Live Location](images/tracking.png)