Introduction

For the Robothon 2023 Challenge, the vision prepossessing required three different tasks:

• localization of the board;
• reading the M5 screen for the slider setting.
• reading the multimeter screen for BYOD part of the challenge has been implemented in this package.

Camera launch

The camera used for this task is realsense D435i, attached to the end effector of the Franka Emika arm.

roslaunch realsense2_camera rs_camera.launch camera:=d435_camera

Slider triangles detection

By detecting the screen and the indicators of the current slider position and the desired position and scaling the distances to the relative position command for the slider on the slot, we could accomplish this task.

rosrun hrii_board_localization screen_read.py

dependencies:

• OpenCV_Version==4.7.0
• rospy
• hrii_robothon_msg

Voltage measurement

rosrun hrii_board_localization text_read.py

dependencies:

• OpenCV_Version==4.7.0
• rospy
• hrii_robothon_msg
• google_cloud_vision
• matplotlib
• tkinter
• PIL

Notes

• A google vision API is required to run the task, make sure the credential for the mentioned API is set up correctly.