Toy Autonomous Car

In this project, we

1. build a Toy Car from a robot car kit purchased from Amazon
2. program the Arduino board on the car to make the car drive “on demand” by a Deep Neural Network
3. train a Deep Neural Network to identify US traffic signs
4. build a image processing pipeline that
   • captures frames from a camera stream
   • pre-process it to isolate region of interest
   • apply DNN prediction on ROI to identify US traffic sign
   • send appropriate command to the car to turn-left, turn-right, or stop

Result

Thank you

A hearty Thank You to my partner in crime Min Wang. He got so interested in the project that he actually bought the car, assembled it and helped shoot the final video too. Thank you Min. Your deep interest in the project propelled me to not give up. Against all odds, we got our car project done! Thanks Buddy!

Source Code

Report