Accident Scape

Accident Scape. Where All Lives Matter. We use machine learning and OCR techniques along with sensors to make driving safer at a lower cost. Yawning, Eye Shutting, Vehicle Diagnostic Statistics are tracked.

What Inspired Us?

We were inspired by the exponential rise in road accidents in recent times, we further researched and found out that in 2021, a rise of 18% on last years number of car crashes was observed in the India in the first 2 quarters itself. The highest ever observed in the last 100 years. Most of these deaths are caused by drivers not paying attention or being too tired to drive. 1 in 6 crashes are caused by drunk drivers. Car Crashes are the 8th leading cause of death and account for 2.2% of deaths globally each year.

What We Learned?

We learned about the different ways of facial detection and its challenges. Using this knowledge we developed an accurate method for facial detection and incorporated this method into our solution. We also learned about how accidents are a huge cause of deaths in many countries. Finally we learned about the different limitations of current technologies in preventing accidents as well as the high costs and lack of availability of prominant technologies in this field.

Problems We Faced

We faced multiple problems through the course of developing this prohject, for example the restrictions of hardware and high amount of testing and debugging required. We also faced severe problems with open-source libraries due to the lack and age of documentation and cross compatiblity. In the beggining we faced difficulties regarding the accuracy of our model however with time we were able to rectify this.

How We Built Out Project?

Overview

Smart Accident Prevention Using Raspberry Pi 3B+, OpenCV and Dlib

Libraries Used

• Python 3.9.2
• Dlib v19.23
• CMake
• VirtualEnvironments
• OpenCV-Python
• Espeak
• Scipy
• Numpy
• Imutils

Built With

• dlib
• gps
• obd-ii
• opencv
• pygame
• python
• raspberry-pi