README.md

 

3D Object Detection and Recognition Based on RGBD Images

---

About   |   Features   |   Technologies   |   Requirements   |   Starting   |   License   |   Author

🎯 About

Package for 3D Object Detection and Recognition Based on RGBD Images in Robot Operating System (ROS) Workspace. This is my final project as student in EEPIS/PENS.

🔗 Paper

https://ieeexplore.ieee.org/document/9594034

@INPROCEEDINGS{9594034,
  author={Birri, Ikmalil and Dewantara, Bima Sena Bayu and Pramadihanto, Dadet},
  booktitle={2021 International Electronics Symposium (IES)}, 
  title={3D Object Detection and Recognition based on RGBD Images for Healthcare Robot}, 
  year={2021},
  volume={},
  number={},
  pages={173-178},
  doi={10.1109/IES53407.2021.9594034}}

📹: Publication

✨ Features

✔️ Feature 1; Realtime RGBD Camera using Intel LibRealsense and opencv
✔️ Feature 2; Filter and Segmentation using Remove NaN, passthrough, VoxelGrid downsample, and RANSAC Plane Segmentation
✔️ Feature 3; Detection or Clustering using Euclidean Cluster
✔️ Feature 4; Prediction Artificial Neural Network using FANN Library
✔️ Feature 5; Visualize 3D Bounding Box and label using Visualizer PCL\

🚀 Technologies

The following tools were used in this project:

• PointCloud
• FANN
• LibRealsense
• OpenCV
• ROS

✅ Requirements

Before starting 🏁, you need to have Git, ROS and PCL installed.

🏁 Starting

# make sure inside catkin_ws/src
$ cd catkin_ws/src

# Clone this project
$ git clone https://github.com/Malikmal/object_recognition_pkg

# Access
$ cd ../

# Build the project (workspace)
$ catkin_make

# Prepare datasets
# contains folders of object scenes

# Run training
$ rosrun object_recognition_pkg main_trainning folder_name

# Run Testing offline file mode
$ rosrun object_recognition_pkg main src/data/scene_mug_table.pcd

# Run realtime camera 
$ rostopic
$ rosrun object_recognition_pkg camera_node
$ rosrun object_recognition_pkg main_camera

📝 License

Made with ❤️ by Ikmalil Birri

 

Back to top