Sky/cloud analysis : cloud classification on a large scale whole sky image dataset

• Main Author: Chia, Sharon Yan Ling
• Other Authors: Lee Yee Hui
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/71174
• Institution: Nanyang Technological University

To understand cloud formations in the sky, ground-based cameras are used progressively nowadays. It can be used in a variety of applications such as for military and meteorological purposes. Classification of clouds is usually done manually by looking through the images one at a time. However, this will affect the accuracy due to different peoples’ perspective of the different cloud types and it is extremely time consuming. This research aims to improve on cloud classification by automating this process. In this report, different methods will be used to increase the accuracy of the classification of the cloud images captured by Wide Angle High Resolution Sky Imaging System (WAHRSIS). By starting the test with images in the Singapore Whole sky IMaging CATegories (SWIMCAT) database, the author will be able to identify if the method used is going in the right direction. The database contains images of sky/cloud that is categorized into 5 distinct categories: • Category A (CAT A) – Clear Sky • Category B (CAT B) – Patterned Clouds • Category C (CAT C) – Thick Dark Clouds • Category D (CAT D) – Thick White Clouds • Category E (CAT E) – Veil Clouds