We made this project when we participated Hack x SJTU competition.
Thanks NVIDIA for their technical support
Jin Yongxu
Luo Anqi
Gou Minghao
Wang Tao
Qian Xinxin
We implemented a smart car using Arduino and TX1. The smart car do some movement as following:
- Move forward when the camera detect a car
- Move Back when detect a chair
- Turn left when detect a cat
- Turn right when detect a person
- Turn left then move forward when detect a bird
- Turn right then move forward when detect a bicycle
- Stop immediatily when the car meet an obstacle(using ultrasonic sensor)
- Caffe Deeplearning framework
- Arduino
- ROS
- OpenCV
- car4.c is the code controlling arduino
- main.cpp is the code controlling TX1
You should use ROS to create a function package named ros_caffe
Then create a node named ros_caffe_test
catkin_make
roscore
rosrun rosserial_python serial_node.py /dev/ttyACM0
rosrun ros_caffe ros_caffe_test
Here is the official website of Hackson.