Vehicle Crash Detection
4 months ago
A fully functioning Computer Vision Program (detect crashes), a mobile app (notify contacts)
What is included in the sale ?
Fully functioning code (both computer vision program and mobile app),
What is the technological stack ?Computer Vision (OpenCV), Python, Java Android (Android Studio)
Which parts of the project are incomplete ?
So far, everything works fine. Computer vision programme is able to detect any kind of collision and mobile application sends SMS to the contacts given by the user. Computer vision programme is able to detect severity level as well. Mobile application has a waiting time before it sends SMS to the contacts and the user can change it in the settings. It plays an emergency alarm as well.
Do you have any traffic ?
Why are you selling ?Don't have enough resources to test it in real conditions. This is my final year project. I want someone to use it for the betterment of the society.
How can the future owner turn the project into a business ?
Automobile companies will gain large number of benefits from this project. If this is modified and developed to the industrial level, this will be a better system than existing AACN systems.
Is this software using machine learning? What dataset is it trained on?
Crash detection program uses OpenCV which is computer vision to detect crashes. Programming language is Python. This program monitors passenger movement and inside of the car which gives very effective results in identifying any kind of crashes. To detect passengers it uses a face detection algorithm then it monitors the movement of the passengers by optical flow. Also it calculates background difference and detects whether car met an accident. Using those two factors this programme successfully detects any kind of crash. This is more advance than I can explain here.
How are the computer vision program and mobile app linked? Is the app streaming a video to a server?
Computer vision program is developed to connect with the mobile app via Bluetooth. So far it works very well. There's a feature in the app to view the camera feed as well. If the user wants to adjust the camera connected to the hardware which is the computer vision program is installed, he can do it using the mobile app. As soon as the program detects the crash it sends a message to the mobile app then mobile app triggers an emergency alarm. There's a waiting time before the app sends emergency SMS to the contacts. If the user is fine he is allowed to stop notifying contacts. The waiting time can be changed from the settings and emergency alarm can be enabled/disabled as well.