Control Custom Robot in ROS Gazebo with Object Detection Using Tensorflow

Created a simulated 2-wheel differential drive robot. Defined inertia matrix, sensor attributes and TF relationships between each part using xacro. Imported the robot to an environment in Gazebo. The angular and linear velocity of the robot is then controlled by a cup’s position detected on the camera in a real-world environment using object detection in TensorFlow.
(More details will be added soon)
Links:


skill_icon-img

Other Projects.

A* search algorithm comparison on Manhattan and Euclidean distance

Control Custom Robot in ROS Gazebo with Object Detection in Tensorflow

2D Mapping using Hector SLAM and RPLidar with Raspberry Pi 4

Computer Vision: Uses machine learning CNN DNN to train model and realize object detection

© 2020 Jinbing Ying.