meetdoshi90/360-object-detection

Repo for 360 degree object detection pipeline

🖥️ 360° Machine Vision For Blind LiDAR Based Hat 🖥️

An Machine Vision System with Object Detection and Depth Estimation with Text to Speech for assisting a blind person

Directory

./src/* 
  ├─ src/AbstractProjection.py - # Parent class for other classes. Includes common functionality and conversion to 3D vectors of images.
  ├─ src/CubemapProjection.py - # Six faced cubemap class which converts 3D vectors to Cubemap projection
  ├─ src/EquirectangularProjection.py - # Equirectangular panaromic view converter
  ├─ src/FisheyeProjection.py - # Fisheye Single lens conversion
  └─ src/SideBySideFisheyeProjection.py # Side by side dual fish eye lens converter
./app.py - # Driver instructions for mapping
./yolov5.py - # Detection and testing from pre-trained Yolov5 and Custom trained Yolov5
./start_server.sh - # Load all files and call app.py

Implementation

TBD...

User Guide

Test Commands

Currently commands are set to Image show.

Can be modified to Image save.

Example command for Fisheye to Cubemap:

python app.py --approach Cubemap --input Image --image_path test.png --output_path "front.png right.png back.png left.png top.png bottom.png" --output_width 512 --output_height 512

Example command for Fisheye to Equi:

python app.py --approach Equirectangular --input Image --image_path test.png --output_path "equi.png" --output_width 512 --output_height 512

Authors

👤 Meet Doshi

• GitHub: @meetdoshi90
• LinkedIn: @meetdoshi90

👤 Abhishek Potdar

👤 Aniket Kulkarni

👤 Shafina Charania

👤 Param Shah

👤 Omkar Gavandi

🤝 Contributing

Contributions, issues, and feature requests are welcome!

Feel free to check the issues page.

Show your support

Give a ⭐️ if you like this project!

📝 License

This project is licensed with GNU General Public License.

Connect with me: