Indoor Visual Positioning & Navigation System
Supervisor : Mohamed Atia
Team size: Minimum , Maximum
| CSE | SE | Comm | Biomed | EE | Aero | Special |
|---|---|---|---|---|---|---|
| Yes | Yes | Yes | No | Yes | Yes | No |
Description
This project aims to develop a smartphone-based indoor positioning system that operates without the need for GPS. Using an Intel RealSense camera, students will first map an environment by capturing geolocated images and extracting visual features. These features are stored in a cloud-based database. The user app then captures live images, extracts features from them, matches these features against the stored map, and estimates the user’s location in real-time. The app will provide pathfinding and turn-by-turn navigation, similar to GPS, in outdoor environments.
Objectives
- Develop a pipeline for image capture, feature extraction, and cloud-based feature storage.
- Implement on-device feature extraction and real-time image matching on a smartphone.
- Estimate user location and orientation from matched visual features.
- Integrate navigation functions such as pathfinding and turn-by-turn guidance.
- Evaluate system accuracy, latency, and energy consumption.
- Discuss legal, privacy, economic, and commercialization considerations.
Educational & Training Value
- Exposure to computer vision (feature detection, image retrieval, pose estimation).
- Embedded/edge AI optimization on smartphones.
- Cloud database management for visual maps.
- Systems integration: sensing, perception, and navigation.
- Consideration of privacy, regulatory, and commercialization aspects.
Expected Outcomes
- A working smartphone app that provides real-time indoor positioning and navigation.
- A cloud-based visual map database for a test building.
- Demonstration of navigation within a mapped environment.
- A final report analyzing technical performance, user experience, and commercialization.
- Analysis of Accuracy, Latency, Energy consumption, Usability, and Privacy
Prerequisites:
- Computer vision (OpenCV, feature extraction, Edge AI/ML, SLAM basics)
- Mobile programming (Android/iOS app development)
- Cloud database systems (feature storage & retrieval)
- Pathfinding algorithms (graph search, shortest path)
- Knowledge of privacy, legal, and ethical issues in data collection