Volume 16, Issue 2 (8-2022)
مرتع 2022, 16(2): 359-378 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Aghababaei M, Ebrahimi A, Naghipour A A, Asadi E. Development of a Google Earth Image's visual Interpretation Protocol to Determine Plant Ecological Units of the Semi-Steppe Regions. مرتع 2022; 16 (2) :359-378
URL: http://rangelandsrm.ir/article-1-1118-en.html
URL: http://rangelandsrm.ir/article-1-1118-en.html
Department of Range and Watershed Management, Faculty of Natural Resources and Earth Sciences, Shahrekord University, Shahrekord
Abstract: (1765 Views)
Background and objectives: Google Earth images are a valuable resource for understanding and studying the natural area's ecology due to their high spatial resolution. Considering that these images have been available for many years, in our country, these valuable data have not been used enough, especially in order to study vegetation and optimal natural areas management. So, the requirement to standardize the utilization of these images seems necessary to study plant ecological units as a basis for natural resource management. In this study, a protocol for google earth images visual interpretation has been developed to extract plant ecological units using six key indicators of shape, texture, color, tone, pattern, and shadow.
Methodology: Related Google Earth images released in 2018 were used. Using the Offline Map Maker software, images related to the study area from google earth with the required maximum magnification (magnification -20) were used. Six key visual interpretation indicators of images were used to prepare a protocol for the plant ecological unit’s separation. In order to have sufficient repetition in defining and applying these indicators, three experts were used. These three experts, having the study area google earth images and based on this protocol, performed plant ecological units’ visual interpretation and separation in GIS10.5 Arc software. Finally, three digital maps were prepared. In order to evaluate the developed protocol, a ground control map was prepared by field visit and taking 200 observation points of the mentioned units. After preparing the plant ecological units maps by each expert, error matrix analysis was used for accuracy analysis and maps validation. So that using the ground reality map obtained from the ground control points for each map, the amount of producer accuracy, user accuracy, kappa index of agreement, and overall accuracy were calculated separately.
Results: According to the key indicators of visual interpretation defined in this protocol, each expert prepared plant ecological unit’s maps. The error matrix analysis results show that the producer accuracy varied between 69 to 96%, but in most units, they had a producer accuracy of more than 80%. In addition, most plant ecological units with one dominant species had higher producer accuracy compared to units with two species predominance. Regarding user accuracy, the accuracy varies between 63 and 96%. The second and third ecological units had 96% user accuracy. Kappa index of agreement also represents an agreement between 68 and 95% for the ground reality map and the maps of the classification (visual interpretation) of experts. While the seventh plant ecological unit agreement is 68%, the first and fourth units 95% agree with the ground reality map. For plant ecological units with two dominant species, in addition to producer accuracy, user accuracy and kappa index of agreement have also decreased. The overall accuracy of the maps produced by these three experts includes 93%, 80%, and 90%, respectively, which have a good overall accuracy, which shows the very high accuracy and spatial resolution of google earth images.
Conclusion: The results showed that there is a significant agreement between experts in the visual interpretation of google earth images. Therefore, plant ecological units could be separated and map prepared with relatively high accuracy. By developing such protocol using key elements introduced, beginner and inexperienced experts can develop maps with acceptable accuracy for better management and different ecosystems organization. However, according to this study's findings, it is recommended that this be used more for units that are structurally different to provide more acceptable accuracy maps. Therefore, it is recommended in areas where the vegetation composition is mostly dominated by individual plant species, and the use of this protocol can lead to a more acceptable accurate visual interpretation of google earth images.
Methodology: Related Google Earth images released in 2018 were used. Using the Offline Map Maker software, images related to the study area from google earth with the required maximum magnification (magnification -20) were used. Six key visual interpretation indicators of images were used to prepare a protocol for the plant ecological unit’s separation. In order to have sufficient repetition in defining and applying these indicators, three experts were used. These three experts, having the study area google earth images and based on this protocol, performed plant ecological units’ visual interpretation and separation in GIS10.5 Arc software. Finally, three digital maps were prepared. In order to evaluate the developed protocol, a ground control map was prepared by field visit and taking 200 observation points of the mentioned units. After preparing the plant ecological units maps by each expert, error matrix analysis was used for accuracy analysis and maps validation. So that using the ground reality map obtained from the ground control points for each map, the amount of producer accuracy, user accuracy, kappa index of agreement, and overall accuracy were calculated separately.
Results: According to the key indicators of visual interpretation defined in this protocol, each expert prepared plant ecological unit’s maps. The error matrix analysis results show that the producer accuracy varied between 69 to 96%, but in most units, they had a producer accuracy of more than 80%. In addition, most plant ecological units with one dominant species had higher producer accuracy compared to units with two species predominance. Regarding user accuracy, the accuracy varies between 63 and 96%. The second and third ecological units had 96% user accuracy. Kappa index of agreement also represents an agreement between 68 and 95% for the ground reality map and the maps of the classification (visual interpretation) of experts. While the seventh plant ecological unit agreement is 68%, the first and fourth units 95% agree with the ground reality map. For plant ecological units with two dominant species, in addition to producer accuracy, user accuracy and kappa index of agreement have also decreased. The overall accuracy of the maps produced by these three experts includes 93%, 80%, and 90%, respectively, which have a good overall accuracy, which shows the very high accuracy and spatial resolution of google earth images.
Conclusion: The results showed that there is a significant agreement between experts in the visual interpretation of google earth images. Therefore, plant ecological units could be separated and map prepared with relatively high accuracy. By developing such protocol using key elements introduced, beginner and inexperienced experts can develop maps with acceptable accuracy for better management and different ecosystems organization. However, according to this study's findings, it is recommended that this be used more for units that are structurally different to provide more acceptable accuracy maps. Therefore, it is recommended in areas where the vegetation composition is mostly dominated by individual plant species, and the use of this protocol can lead to a more acceptable accurate visual interpretation of google earth images.
Keywords: Key indicators of visual image interpretation, Rangeland, Land cover map, Overall accuracy
Type of Study: Research |
Subject:
Special
Received: 2022/01/20 | Accepted: 2022/04/24 | Published: 2022/08/1
Received: 2022/01/20 | Accepted: 2022/04/24 | Published: 2022/08/1
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
