A visual-based tourism assistant system (Server and algorithm part)

• Main Author: Chen, Liyang
• Other Authors: Cai Jianfei
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/66892
• Institution: Nanyang Technological University

This project is a joint Final Year Project conducted by Zhou Xinzi and me. It aims to build a visual-based tourism assistant system. The system is designed for mobile phones or wearable devices, such as see-through glasses, and the core functionalities are locating the user using video stream from device camera and marking surrounding objects. We investigated several types of image features and image-based localization algorithms. On top of researches by Rapid-Rich Object Search (ROSE) Lab and researches on image- based indoor localization by Tao Qingyi in her final year project, the system uses a client-server model and Compact Descriptor for Visual Search (CDVS) solution from ROSE Lab. We explored various system architectures and object recognition strategies. We took 1326 images at NTU campus as our dataset for testing and archived good accuracy for localization and object recognition. We also profiled our system on aspects of speed, memory and network, and the results show that our system is good for the mobile platform. In addition, the system provides good user experience.