Student research opportunities
Shape and other properties of transparent surfaces
Project Code: CECS_950
This project is available at the following levels:
PhD
Please note that this project is only for higher degree (postgraduate) applicants.
Keywords:
shape reconstruction, transparent surfaces, refraction, refractive index, image processing, computer vision
Supervisors:
Dr Cong Phuoc HuynhDr Antonio Robles-kelly
Outline:
The refraction of light through transparent surfaces gives rise to multiple clues related to the instrinsic properties of the material and the scene background. Even with state-of-the-art sensors, such as laser rangers and time of flight cameras, acquiring the 3D representation of transparent surfaces remains a challenging task.
Despite the underlying challenges, an analysis of transparency through visual inputs (images and videos) yields many potential applications. For example, the shape and refractive index of the transparent object could be recovered for graphics and recognition tasks. In another
scenario, an accurate model of transparency could also help rectifying the background behind transparent surfaces, enhancing the visibility of scenes captured under-water.
Goals of this project
The project aims to recover the shape and material properties of transparent surfaces. A by-product of this study is also the estimation of the background geometry (depth) and apperance (colour). The acquisition of visual inputs will be performed with a flexible
range of imaging modalities such as multispectral, infrared, polarisation and plenoptic cameras.
Requirements/Prerequisites
This PhD project is suitable for candidates with strong background in mathematics, physics, computer science and engineering and a first class Honours degree in these related areas or equivalent. Good foundations in mathematics (linear algebra, calculus and optimisation) and physics (optics) are essential to the success of the candidature. In addition, candidates should have good software design and programming skills and experience with C++ and Matlab.
Student Gain
The student will gain a hands-on experience working with both advanced imaging equipment as well as imaging processing and computer vision techniques under the supervision of motivated and experienced researchers in the related areas. In case the research results show potential practicality, the student will be involved in patenting and commercialisation activities.
Background Literature
1. Z.Chen, K. Wong, Y. Matsushita and X. Zhu. Depth from Refraction Using a Transparent Medium with Unknown Pose and Refractive Index. International Journal of Computer Vision 2013
2. A. Levin and Y. Weiss. User Assisted Separation of Reflections from a Single Image Using a Sparsity Prior. IEEE Transactions on Pattern Analysis and Machine Intelligence 2007.
