Search published articles


Showing 2 results for Object-Oriented Classification

Khadijeh Haji, Abazar Esmali-Ouri, Raoof Mostafazadeh, Dr Habib Nazarnejad,
Volume 22, Issue 66 (9-2022)
Abstract

Also, because of human activities and natural phenomena, the face of the earth has always undergone a change. Therefore, for optimal management of natural areas, awareness of the ratio of land cover/land use changes is a necessity. Therefore, extraction of land use maps as the most important goal in the management of the natural resource can be considered. The purpose of the present study was to evaluate land cover/ land use changes at the Rozechai Watershed during the period of 30-years 1985-2015 using Landsat 5 and Landsat 7 satellite imageries such as TM and ETM+ sensors; plus, land use maps were prepared using TerrSet software and object-oriented classification in 1985 and 2000 years. As well as the land use map of procurement by the geographical organization in 2015 has been used. The results show that rangelands level has the highest percentage among all land use types during the period of 30 years, but between 1985 and 2000, and 2000 to 2015, the level of rangelands has a decreasing trend indicating the destruction trend in the region of the replacement of moderate- poor rangelands and good rangeland by dry farming. Also, the tables of obtained from the error matrix indicate that the observed values in the diameter of the error matrix are much larger than the values outside the diameters. Thus, the overall accuracy for the years 1985, 2000, and 2015 were 97, 90 and 96 percent, and The values of Kappa index were 91%, 84% and 94% respectively, indicating a high degree of accuracy in the object-oriented approach to classification.

Mahrookh Ghazayi, Nazfar Aghazadeh, Ehsan Ghaleh, Elhameh Ebaddyy,
Volume 25, Issue 79 (12-2025)
Abstract

The depletion of surface water resources has necessitated uncontrolled groundwater abstraction in various regions worldwide, resulting in substantial reductions in groundwater table levels. As populations continue to expand, the extraction of these essential resources has intensified, posing a significant threat to natural reserves. This study aims to monitor groundwater levels through the analysis of satellite imagery and to investigate the correlation between these levels and land use patterns. To accomplish this objective, relevant satellite images were acquired and subjected to appropriate pre-processing. An object-oriented methodology was employed to generate land use classification maps for two distinct years, alongside a land use change map covering a fifteen-year period from 2000 to 2015. Moreover, groundwater level maps for the study area were produced for both years utilizing the Gaussian method, recognized as the most accurate approach. The findings indicate a robust and significant relationship between land use and groundwater levels, revealing that areas with higher vegetation exhibit lower groundwater levels compared to other regions. This phenomenon can be attributed to the hydrological dynamics that facilitate the movement of water from higher potential zones to these areas. Additionally, irrigated agricultural practices demonstrated the most pronounced average decline in water levels relative to other land uses, underscoring the excessive reliance on groundwater for irrigation in the study area. The results further illustrate that the conventional kriging method with Gaussian variance surpasses other techniques in estimating groundwater table depths across both statistical periods. Analysis through conventional kriging reveals a general decline in groundwater levels throughout the majority of the plain during the study period, with a maximum decrease of 40 meters and an average reduction of 15 meters.


Page 1 from 1     

Creative Commons License
This work is licensed under a Creative Commons — Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)