Cite Article

Efficient Supervised Features Learning for Remote Sensing Image Classification

Choose citation format

BibTeX

@article{IJASEIT11272,
   author = {Sarah Qahtan Mohammed Salih and Abdul Sattar Arif Khammas and Ramlan Mahmod},
   title = {Efficient Supervised Features Learning for Remote Sensing Image Classification},
   journal = {International Journal on Advanced Science, Engineering and Information Technology},
   volume = {11},
   number = {2},
   year = {2021},
   pages = {549--558},
   keywords = {Convolutional neural networks; remote sensing; image classification; feature extraction; pre-trained; fine-tuned.},
   abstract = {

The features extracted from the fully connected (FC) layers of a convolutional neural network (ConvNet or CNN) can provide accurate classification results as long as the labelled datasets are large enough. On the other end, high accuracy remote sensing image (RSI) classification is demanded various implementations such as urban planning, environmental monitoring, and geographic image retrieval. Many studies have been presented in this domain; however, satisfactory classification accuracy is yet to be achieved. In this study, the proposed method of fine-tuning the pre-trained ConvNets (GoogleNet, VGG16, and ResNet50) on RSI, extracting features from the last fine-tuned FC layer of these networks and reprocess the extracted features for classification by SVM, produced high classification accuracy. Extensive experiments have been conducted on three RSI datasets: the NWPU, AID, and PatternNet. Comparative results over the selected datasets demonstrate that our method considerably outperforms the state-of-the-art best-stated results. Also, the overall accuracy (OA) and confusion matrix report quantitative evaluation. Our best outcomes from the first part were 99.54%, 94.60%, and 94.83% on the PatternNet, NWPU, and AID datasets, respectively, achieved by fine-tuned ResNet50. Moreover, the best classification accuracies with training ratios 20% and 50% on the AID dataset, 10% and 20% on the NWPU dataset, and the 10%, 20%, 50% and 80% on PatternNet dataset were 95.72%, 97.53%, 96.19%, 96.85%, 99.60%, 99.56%, 99.75% and 99.80% respectively. The classification performance of each class was estimated using a confusion matrix for the three datasets.

},    issn = {2088-5334},    publisher = {INSIGHT - Indonesian Society for Knowledge and Human Development},    url = {http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=11272},    doi = {10.18517/ijaseit.11.2.11272} }

EndNote

%A Salih, Sarah Qahtan Mohammed
%A Khammas, Abdul Sattar Arif
%A Mahmod, Ramlan
%D 2021
%T Efficient Supervised Features Learning for Remote Sensing Image Classification
%B 2021
%9 Convolutional neural networks; remote sensing; image classification; feature extraction; pre-trained; fine-tuned.
%! Efficient Supervised Features Learning for Remote Sensing Image Classification
%K Convolutional neural networks; remote sensing; image classification; feature extraction; pre-trained; fine-tuned.
%X 

The features extracted from the fully connected (FC) layers of a convolutional neural network (ConvNet or CNN) can provide accurate classification results as long as the labelled datasets are large enough. On the other end, high accuracy remote sensing image (RSI) classification is demanded various implementations such as urban planning, environmental monitoring, and geographic image retrieval. Many studies have been presented in this domain; however, satisfactory classification accuracy is yet to be achieved. In this study, the proposed method of fine-tuning the pre-trained ConvNets (GoogleNet, VGG16, and ResNet50) on RSI, extracting features from the last fine-tuned FC layer of these networks and reprocess the extracted features for classification by SVM, produced high classification accuracy. Extensive experiments have been conducted on three RSI datasets: the NWPU, AID, and PatternNet. Comparative results over the selected datasets demonstrate that our method considerably outperforms the state-of-the-art best-stated results. Also, the overall accuracy (OA) and confusion matrix report quantitative evaluation. Our best outcomes from the first part were 99.54%, 94.60%, and 94.83% on the PatternNet, NWPU, and AID datasets, respectively, achieved by fine-tuned ResNet50. Moreover, the best classification accuracies with training ratios 20% and 50% on the AID dataset, 10% and 20% on the NWPU dataset, and the 10%, 20%, 50% and 80% on PatternNet dataset were 95.72%, 97.53%, 96.19%, 96.85%, 99.60%, 99.56%, 99.75% and 99.80% respectively. The classification performance of each class was estimated using a confusion matrix for the three datasets.

%U http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=11272 %R doi:10.18517/ijaseit.11.2.11272 %J International Journal on Advanced Science, Engineering and Information Technology %V 11 %N 2 %@ 2088-5334

IEEE

Sarah Qahtan Mohammed Salih,Abdul Sattar Arif Khammas and Ramlan Mahmod,"Efficient Supervised Features Learning for Remote Sensing Image Classification," International Journal on Advanced Science, Engineering and Information Technology, vol. 11, no. 2, pp. 549-558, 2021. [Online]. Available: http://dx.doi.org/10.18517/ijaseit.11.2.11272.

RefMan/ProCite (RIS)

TY  - JOUR
AU  - Salih, Sarah Qahtan Mohammed
AU  - Khammas, Abdul Sattar Arif
AU  - Mahmod, Ramlan
PY  - 2021
TI  - Efficient Supervised Features Learning for Remote Sensing Image Classification
JF  - International Journal on Advanced Science, Engineering and Information Technology; Vol. 11 (2021) No. 2
Y2  - 2021
SP  - 549
EP  - 558
SN  - 2088-5334
PB  - INSIGHT - Indonesian Society for Knowledge and Human Development
KW  - Convolutional neural networks; remote sensing; image classification; feature extraction; pre-trained; fine-tuned.
N2  - 

The features extracted from the fully connected (FC) layers of a convolutional neural network (ConvNet or CNN) can provide accurate classification results as long as the labelled datasets are large enough. On the other end, high accuracy remote sensing image (RSI) classification is demanded various implementations such as urban planning, environmental monitoring, and geographic image retrieval. Many studies have been presented in this domain; however, satisfactory classification accuracy is yet to be achieved. In this study, the proposed method of fine-tuning the pre-trained ConvNets (GoogleNet, VGG16, and ResNet50) on RSI, extracting features from the last fine-tuned FC layer of these networks and reprocess the extracted features for classification by SVM, produced high classification accuracy. Extensive experiments have been conducted on three RSI datasets: the NWPU, AID, and PatternNet. Comparative results over the selected datasets demonstrate that our method considerably outperforms the state-of-the-art best-stated results. Also, the overall accuracy (OA) and confusion matrix report quantitative evaluation. Our best outcomes from the first part were 99.54%, 94.60%, and 94.83% on the PatternNet, NWPU, and AID datasets, respectively, achieved by fine-tuned ResNet50. Moreover, the best classification accuracies with training ratios 20% and 50% on the AID dataset, 10% and 20% on the NWPU dataset, and the 10%, 20%, 50% and 80% on PatternNet dataset were 95.72%, 97.53%, 96.19%, 96.85%, 99.60%, 99.56%, 99.75% and 99.80% respectively. The classification performance of each class was estimated using a confusion matrix for the three datasets.

UR - http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=11272 DO - 10.18517/ijaseit.11.2.11272

RefWorks

RT Journal Article
ID 11272
A1 Salih, Sarah Qahtan Mohammed
A1 Khammas, Abdul Sattar Arif
A1 Mahmod, Ramlan
T1 Efficient Supervised Features Learning for Remote Sensing Image Classification
JF International Journal on Advanced Science, Engineering and Information Technology
VO 11
IS 2
YR 2021
SP 549
OP 558
SN 2088-5334
PB INSIGHT - Indonesian Society for Knowledge and Human Development
K1 Convolutional neural networks; remote sensing; image classification; feature extraction; pre-trained; fine-tuned.
AB 

The features extracted from the fully connected (FC) layers of a convolutional neural network (ConvNet or CNN) can provide accurate classification results as long as the labelled datasets are large enough. On the other end, high accuracy remote sensing image (RSI) classification is demanded various implementations such as urban planning, environmental monitoring, and geographic image retrieval. Many studies have been presented in this domain; however, satisfactory classification accuracy is yet to be achieved. In this study, the proposed method of fine-tuning the pre-trained ConvNets (GoogleNet, VGG16, and ResNet50) on RSI, extracting features from the last fine-tuned FC layer of these networks and reprocess the extracted features for classification by SVM, produced high classification accuracy. Extensive experiments have been conducted on three RSI datasets: the NWPU, AID, and PatternNet. Comparative results over the selected datasets demonstrate that our method considerably outperforms the state-of-the-art best-stated results. Also, the overall accuracy (OA) and confusion matrix report quantitative evaluation. Our best outcomes from the first part were 99.54%, 94.60%, and 94.83% on the PatternNet, NWPU, and AID datasets, respectively, achieved by fine-tuned ResNet50. Moreover, the best classification accuracies with training ratios 20% and 50% on the AID dataset, 10% and 20% on the NWPU dataset, and the 10%, 20%, 50% and 80% on PatternNet dataset were 95.72%, 97.53%, 96.19%, 96.85%, 99.60%, 99.56%, 99.75% and 99.80% respectively. The classification performance of each class was estimated using a confusion matrix for the three datasets.

LK http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=11272 DO - 10.18517/ijaseit.11.2.11272