Student research opportunities
3D modeling of plants
Project Code: CECS_1172
This project is available at the following levels:
CS single semester, Engn R&D, Honours, Masters
Keywords:
3D computer vision, 3D scanning, plants, non-destructive measurement
Supervisor:
Dr Chuong NguyenOutline:
Human is facing new challenges in this new century:
- Growing population and constrained resources. UN estimated that we need an increase of 70% of food production by 2050. This increase needs to be achieved with current available fixed resources such as agricultural land area, energy supply, fresh water supply, the use and disposal of fertilizers, pesticides and other by products.
- Climate change. This global change in the climate and weather conditions is making traditional food production ineffective and changing ecosystem. Drought and salinity significantly reduces productivity.
- Invasive pests and diseases. As globalisation remove boundaries between regions and nations, pest and diseases can easily transmit from one place to another. Invasive pest can destroy major crops and pose major risk to important agricultural industries.
- Limited progress in understanding biological system. Traditional methods of improving our knowledge of nature do not allow us to keep up with the new demands and challenges. With current progress in crop productivity, there will not be enough food for everyone by 2050.
To make a breakthrough, new research tools and methodologies are needed.
Goals of this project
A 3D plant scanner and software pipeline will be developed to automatically scan and construct complete 3D models of plants. The 3D models are accurate enough to extract anatomical characteristics or growth patterns of the plants.
Requirements/Prerequisites
Computer Vision (ENGN4528), C/C++ Matlab or Python programming.
Student Gain
Apply computer vision techniques to solve a real-world interdisciplinary problem. Learn camera calibration and 3D reconstruction techniques. Work with specialised camera system. Create beautiful 3D plant models.
