Tools: SLAM BRISK ROS Mobile platform

MLF6110

Code

In this project, my teammates and I developed a mobile robot to search for people’s lost properties in a confined space. The motivation of this project is that we often see people posting in campus online groups looking for their lost properties, so we try to solve this issue using a mobile robot.

The robot is equipped with RealSense and RPLIDAR-A1. The user can upload a picture of their lost properties on a remote computer. The RealSense onboard would do template matching to find the object in the room. If it finds it, it would record the coordinate of the object and back to the origin. The user can then type in the object wish to find; led by the robot, the user can find his or her lost property.

In this project, we implement BRISK algorithm as feature points detector and descriptor, brute-force hamming as feature points matcher and RANSAC as homography method. The SLAM algorithm we use is GMapping.