Please use this identifier to cite or link to this item:
|Title:||3D reconstruction of cloud cells||Authors:||Ng, Felicia Ai Jing||Keywords:||DRNTU::Engineering||Issue Date:||2016||Abstract:||Ground-based sky cameras are increasingly used now-a-days to understand cloud formation analysis in the atmosphere. Such cloud analysis has varied applications in aviation industry, solar and renewable energy predictions and cloud attenuation analysis. In this report, the author is interested to perform a 3D cloud reconstruction using a pair of ground-based sky cameras. Conventional feature matching approaches in computer vision community could not be directly applied to atmospheric clouds, which often seen as featureless, and that may pose as a challenge in such computer vision analysis. This project aims to identify the most effective feature matching algorithm that could maximise the performance of 3D cloud cells reconstruction. Also, parameters which affect the respective feature matching algorithm performances would be covered. Putting particular focus on various types of clouds, in order to improve the feature point matching performance and efficiency, experimental results on cumulus and dark stratocumulus clouds have reflected that the proposed algorithm of combining SURF feature point matching and adaptive histogram equalization contrast technique, have resulted in stronger robustness in a variety of complex image cases, such as dark clouds and stratocumulus clouds for 3D cloud cell reconstruction.||URI:||http://hdl.handle.net/10356/67115||Rights:||Nanyang Technological University||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.