Novel Underwater Camera for Marine Life Surveying
Project Description
Omnidirectional camera is a cost-effective and information-rich sensor, which is highly suitable for many marine applications and the ocean scientific community, encompassing several domains such as augmented reality, mapping, motion estimation, visual surveillance, and simultaneous localization and mapping. This project aims to design and construct such a high-quality 360-degree real-time streaming camera system for extended periods for underwater applications.
Prefer: Third year and final year student, students who have had experience with robotics, students who want to pursue post-graduate, final year students for final year projects.
Prefer: Third year and final year student, students who have had experience with robotics, students who want to pursue post-graduate, final year students for final year projects.
Supervisor
YEUNG, Sai Kit
Quota
5
Course type
UROP1000
UROP1100
UROP2100
UROP3100
UROP3200
UROP4100
Applicant's Roles
a. Software
- Preprocess data for video stitching: data collection, data generation. Learn computer vision algorithms, implement image, and video stitching algorithms.
b. Hardware
- Design electronics system of cameras, design PCB circuit to minimize the space of the whole system, and calculate power consumption, heat of the system.
c. Mechanics
- Design housing to contain all cameras, a mini-computer, cables, electronics circuit which can overcome overheating issue, stand under high-pressure of deep sea and avoid optical issues.
- Preprocess data for video stitching: data collection, data generation. Learn computer vision algorithms, implement image, and video stitching algorithms.
b. Hardware
- Design electronics system of cameras, design PCB circuit to minimize the space of the whole system, and calculate power consumption, heat of the system.
c. Mechanics
- Design housing to contain all cameras, a mini-computer, cables, electronics circuit which can overcome overheating issue, stand under high-pressure of deep sea and avoid optical issues.
Applicant's Learning Objectives
a. Software:
- Learn computer vision knowledge, deep learning, processing data for training models.
b. Hardware:
- Learn and apply embedded systems in a real project.
c. Mechanics:
- Apply theory to real application.
- Learn computer vision knowledge, deep learning, processing data for training models.
b. Hardware:
- Learn and apply embedded systems in a real project.
c. Mechanics:
- Apply theory to real application.
Complexity of the project
Moderate