After a car accident, the prospect of dealing with insurance adjusters and claim forms can be more stressful than the accident itself. Our Claim 0x system would save car insurers time and money while saving the customer to hassle of the claim process
What it does
Collected a great number of pictures from Google and some from our own parking lot to train a data set using Google Tensorflow to determine car damage. Additionally, we used the MPU 6050 accelerometer/gyroscope sensor to determine car accidents based on G force exerted.
How we built it
We collected pictures from Google and some from our own parking lot to train a data set using Google Tensorflow. The MPU 6050 is controlled by an microcontroller and has been optimized and calibrated to triggered when an outstanding G-Force is exerted in all lateral directions.(Much higher than super cars such as the Bugatti).
Challenges we ran into
Windows update! Crashed our pc and held our code hostage for hours.
Accomplishments that we're proud of
Very accurate prediction for whether a car is wrecked or not. We have tested this random pictures numerous times and gotten great results. We're also proud that a portion of our data set for pictures of cars came from us taking pictures of cars in our building.
What we learned
Don't trust Windows, back up code.