anirudh1313 / self_driving_car_earthquake

Autopilot System Design for Natural Disaster

Introduction

This project aims at designing an Autonomous Car that can traverse to the destination at the time of a natural disaster, when the roads and surroundings are destroyed. A 2-D LiDAR and a Stereo Camera are the two sensors used to detect obstacles. Even though LiDAR is fast and accurate, it fails to detect glass, smoke, fog and objects below its 2d scanning plane. Stereo camera is used primarily to detect obstacles under the height at which Lidar is placed. FCN-AlexNet model was trained to segment cracks on the roads, vegetation, water, people, cars, buildings, rocks, and other objects in the surrounding area. The sensor fusion of LiDAR and Camera helps in obstacle detection in all the scenarios, as Lidar cant detect transparent materials such as glass, and fire etc.

Advisor Prof. Hyeran Jeon

Email ID : [email protected]

Web page : https://sites.google.com/a/sjsu.edu/hyeran-jeon/

Team members

1. Katamreddy, Krishna Sai Anirudh

   Email ID : [email protected]

   LinkedIn : https://linkedin.com/in/krishna-sai-anirudh-katamreddy-92a46193/

2. Patel, Doyal

   Email ID : [email protected]

   LinkedIn : https://www.linkedin.com/in/doyal-patel-747a80127/

3. Patel, Kathan

   Email ID : [email protected]

   LinkedIn : https://www.linkedin.com/in/kathan15/

4. Visharia, Raj

   Email ID : [email protected]

   LinkedIn : https://www.linkedin.com/in/rajvisharia/