chess_baxter_human

code allowing Baxter the robot to play chess against a human opponent

Step 0: Clone the repo

git clone https://github.com/kevinkengne/chess_baxter_human.git

Step 1: Install matterport Mask-RCNN inside chess_baxter_human

git submodule update --init --recursive

Step 2: Trained model setup

Download the trained model:

https://drive.google.com/open?id=1VNXWBpbm8BN6mraHUzSQPguF2HFidSN9

copy it inside chess_baxter_human/Mask_RCNN/logs/

Step 3: Install the requirements

pip install -r requirements.txt

Step 4: Setup the usb cam node

rosparam set usb_cam/image_width 960

rosparam set usb_cam/image_height 720

rosrun usb_cam usb_cam_node

Step 5: Set up moveit

rosrun baxter_interface joint_trajectory_action_server.py

roslaunch baxter_moveit_config baxter_grippers.launch

Step 6: Launch the program

1. Fill out scripts/measurements.py using print_joint_angles() from robot_pick_place.py

rosrun chess_baxter_human robot_pick_place

2. Start the node that will record images

rosrun chess_baxter_human image_record

4. Start the node that will get images ready for board state detection

rosrun chess_baxter_human image_to_squares

5. Start the node that does the image inference

rosrun chess_baxter_human image_analysis

6. Start the node that generates chess moves

rosrun chess_baxter_human chess_moves_generator

7. Start the node responsible for the robotic manipulation

rosrun chess_baxter_human robot_pick_place