Meteorology visibility estimation by using multi-support vector regression method

Wai-Lun LO, Meimei ZHU, Hong FU

Research output: Contribution to journalArticlespeer-review

2 Citations (Scopus)

Abstract

Meteorological visibility measures the transparency of the atmosphere or air and it provides important information for road, flight and sea transportation safety. Problem of pollution can also affect the visibility of a certain area. Measurement and estimation of visibility is a challenging and complex problem as visibility is affected by various factors such as dust, smoke, fog and haze. Traditional digital image-based approach for visibility estimation involve applications of the meteorology law and mathematical analysis. Digital image-based and machine learning approach can be one of the solutions to this complex problem. In this paper, we propose an intelligent digital method for visibility estimation. Effective regions are first extracted from the digital images and then classified into different classes by using Support Vector Machines (SVM). Multi-Supported Vector Regression (MSVR) models are used to predict the meteorological visibility by using the image features values generated by VGG Neural Network. SVR machine learning method is used for model training and the resulting system can be used for meteorological visibility estimation. Copyright © 2020 by the authors.
Original languageEnglish
Pages (from-to)40-47
JournalJournal of Advances in Information Technology
Volume11
Issue number2
DOIs
Publication statusPublished - May 2020

Citation

Lo, W. L., Zhu, M., & Fu, H. (2020). Meteorology visibility estimation by using multi-support vector regression method. Journal of Advances in Information Technology, 11(2), 40-47. doi: 10.12720/jait.11.2.40-47

Keywords

  • Meteorology visibility
  • Weather photo
  • Deep learning
  • Feature extraction
  • Support vector regression

Fingerprint Dive into the research topics of 'Meteorology visibility estimation by using multi-support vector regression method'. Together they form a unique fingerprint.