vision_robot_ros

Installation

1. Install ROS Noetic, OpenCV, find-object, Qt, Eigen3, roslib and pymongo.

2. Clone vision_robot_ros into your ROS workspace.

git clone https://github.com/KorawitGems/vision_robot_ros.git

Run

1. Feature Detection

This node publishes a Marker Array containing 3D position objects detected using feature detection from find-object.

https://github.com/introlab/find-object

Requirements:

1.Set the object picture name to name_number.png, such as cat_1.png, dog_2.png. 2.Set the picture folder path in the launch file before running.

roslaunch vision_robot_ros feature_detection_3d.launch

2. Object Detection

This node publishes a Marker Array containing 3D position objects detected using a Deep Learning model with OpenCV DNN.

Requirements:

1.Train an object detection model using a deep learning platform such as TensorFlow. 2.Replace the trained model path in the launch file before running.

roslaunch vision_robot_ros rgbd_object_detect_node.launch

3. Object Database

Saved object's position to mongodb.

roslaunch vision_robot_ros object_mongodb.launch