Evaluation of Object-Based and Pixel-Based Technique for Extracting Snow Cover Surface Using Landsat 8 Satellite Images (Case Study Damavand Mountain Range)

Document Type : Original Article

Authors

1 Water Resources Engineering Department, Faculty of Civil Engineering, Tabriz University, Tabriz, Iran

2 Department of civil engineering, Mohaghegh Ardabili university, Ardabil, Iran

3 Department of civil engineering, Islamic Azad university of Ardabil branch, Ardabil, Iran

Abstract

Accurate snow cover extraction is crucial in water resources management, particularly in regions where snowfall contributes to atmospheric precipitation. However, it poses challenges in mountainous areas due to limited accessibility, diverse topographic and physiographic features, and insufficient meteorological stations. To overcome these limitations, remote sensing, which offers multiple advantages like providing information at different scales, extensive coverage, and cost-effectiveness, is employed to assess various snow cover extraction methods in mountainous regions. This study aimed to assess the effectiveness of object-oriented and pixel-based techniques in extracting snow cover using Landsat8 satellite imagery. The pixel-based method relies on classifying numerical values of images, while the novel object-oriented approach takes into account not only numerical images but also background information, texture, and content for classification. The SAM classification method, a pixel-based technique, and object-oriented classification methods, along with NDSI, NDVI, and LST algorithms, were utilized to process the images. Thematic maps were derived from each classification, and their overall accuracy was evaluated in the post-processing stage. The results revealed that the object-oriented classification method exhibited a general accuracy of 92%, outperforming the pixel-based method, which achieved a general accuracy of 81.6%. This demonstrates that the object-oriented method is more precise in extracting snow cover in the mountainous area of Damavand.

Keywords

Advances in Engineering and Intelligence Systems
Volume 002, Issue 04
December 2023

Files

History

  • Receive Date: 28 November 2023
  • Revise Date: 24 December 2023
  • Accept Date: 25 December 2023

Share

How to cite

Statistics