|Topic:||Localization of a passive mobile carrier|
|Description:||The aim is to develop a camera-based localization system for a mobile carrier pushed by a human. The main application is in a robotized warehouse where to human needs to be localized, so that the robots' trajectories can be recomputed to avoid the getting close to human workers. The system will be based on four algorithms that will have to be combined:
- visual odometry for high speeds,
- local feature matching for low speeds,
- global feature matching for the absolute accuracy
- landmark detection and decoding for the absolute accuracy (this algorithm is specific to the warehouse system, other are general).
|Bibliography:||- A. Heinrich, "An Optical Flow Odometry Sensor Based on the Raspberry Pi Computer". Masters' thesis, 2017.|