Comparison of Deep Learning Algorithms for Image Segmentation on Satellite Images

Hüseyin Acemli; Nida Kumbasar

Makale detayları

Comparison of Deep Learning Algorithms for Image Segmentation on Satellite Images

Yazarlar:

Hüseyin Acemli | Nida Kumbasar

Yayıncı:

Gazi University

Ana Başlıklar:

Artificial Intelligence (Other) | Photogrammetry and Remote Sensing | Remote Sensing | Computational Modelling and Simulation in Earth Sciences

Dergi:

Gazi University Journal of Science Part A: Engineering and Innovation Vol.12 Issue 2, pp.479-502

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001348247

pISSN:

2147-9542

Anahtar kelimeler:

Özet:

 Recent advancements in deep learning have significantly contributed to the development of high spatial resolution HSR land cover mapping. However, the distinct geographic patterns between urban and rural areas have limited the generalizability of deep learning algorithms across these domains. To address this challenge, separate datasets for rural and urban environments have been proposed in the literature, aiming to achieve more reliable results in real-world applications. In this study, we utilize the publicly available LoveDA HSR dataset for model and parameter comparison. Experiments were conducted on two distinct scenarios: rural and urban areas. The combination of the Adam optimizer, Dice loss function, and UNet architecture exhibited the highest performance in both datasets. A weighted average of this combination, based on the number of test samples, was calculated for both groups, yielding a final performance score of 62.14 in terms of mean Intersection over Union IoU.

 Recent advancements in deep learning have significantly contributed to the development of high spatial resolution HSR land cover mapping. However, the distinct geographic patterns between urban and rural areas have limited the generalizability of deep learning algorithms across these domains. To address this challenge, separate datasets for rural and urban environments have been proposed in the literature, aiming to achieve more reliable results in real-world applications. In this study, we utilize the publicly available LoveDA HSR dataset for model and parameter comparison. Experiments were conducted on two distinct scenarios: rural and urban areas. The combination of the Adam optimizer, Dice loss function, and UNet architecture exhibited the highest performance in both datasets. A weighted average of this combination, based on the number of test samples, was calculated for both groups, yielding a final performance score of 62.14 in terms of mean Intersection over Union IoU.

 Recent advancements in deep learning have significantly contributed to the development of high spatial resolution HSR land cover mapping. However, the distinct geographic patterns between urban and rural areas have limited the generalizability of deep learning algorithms across these domains. To address this challenge, separate datasets for rural and urban environments have been proposed in the literature, aiming to achieve more reliable results in real-world applications. In this study, we utilize the publicly available LoveDA HSR dataset for model and parameter comparison. Experiments were conducted on two distinct scenarios: rural and urban areas. The combination of the Adam optimizer, Dice loss function, and UNet architecture exhibited the highest performance in both datasets. A weighted average of this combination, based on the number of test samples, was calculated for both groups, yielding a final performance score of 62.14 in terms of mean Intersection over Union IoU.

 Recent advancements in deep learning have significantly contributed to the development of high spatial resolution HSR land cover mapping. However, the distinct geographic patterns between urban and rural areas have limited the generalizability of deep learning algorithms across these domains. To address this challenge, separate datasets for rural and urban environments have been proposed in the literature, aiming to achieve more reliable results in real-world applications. In this study, we utilize the publicly available LoveDA HSR dataset for model and parameter comparison. Experiments were conducted on two distinct scenarios: rural and urban areas. The combination of the Adam optimizer, Dice loss function, and UNet architecture exhibited the highest performance in both datasets. A weighted average of this combination, based on the number of test samples, was calculated for both groups, yielding a final performance score of 62.14 in terms of mean Intersection over Union IoU.

Makale detayları

Comparison of Deep Learning Algorithms for Image Segmentation on Satellite Images

Yazarlar:

Hüseyin Acemli | Nida Kumbasar

Yayıncı:

Gazi University

Ana Başlıklar:

Artificial Intelligence (Other) | Photogrammetry and Remote Sensing | Remote Sensing | Computational Modelling and Simulation in Earth Sciences

Dergi:

Gazi University Journal of Science Part A: Engineering and Innovation Vol.12 Issue 2, pp.479-502

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001348247

pISSN:

2147-9542

Remote Sensing

High-Spatial Resolution

Semantic Segmentation

Deep Learning

Sayı

Makale detayları

Comparison of Deep Learning Algorithms for Image Segmentation on Satellite Images

Yazarlar: Hüseyin Acemli | Nida Kumbasar

Yayıncı: Gazi University

Ana Başlıklar: Artificial Intelligence (Other) | Photogrammetry and Remote Sensing | Remote Sensing | Computational Modelling and Simulation in Earth Sciences

Dergi: Gazi University Journal of Science Part A: Engineering and Innovation Vol.12 Issue 2, pp.479-502

Belge Tipi: Makale

Yayın Tarihi: 6/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001348247

pISSN: 2147-9542

Remote Sensing

High-Spatial Resolution

Semantic Segmentation

Deep Learning

Sayı

Yazarlar:

Hüseyin Acemli | Nida Kumbasar

Yayıncı:

Gazi University

Ana Başlıklar:

Artificial Intelligence (Other) | Photogrammetry and Remote Sensing | Remote Sensing | Computational Modelling and Simulation in Earth Sciences

Dergi:

Gazi University Journal of Science Part A: Engineering and Innovation Vol.12 Issue 2, pp.479-502

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001348247

pISSN:

2147-9542