Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification
Land use land cover (LULC) classification using satellite images is crucial for land-use inventories and environment modeling. The LULC classification is a difficult task because of the high dimensional feature space, which affects the classification accuracy. This article proposes a dual strategy-b...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10815620/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832582372016521216 |
|---|---|
| author | Vinaykumar Vajjanakurike Nagaraju Ananda Babu Jayachandra Andrzej Stateczny Swathi Holalu Yogesh Raviprakash Madenur Lingaraju Balaji Prabhu Baluvaneralu Veeranna |
| author_facet | Vinaykumar Vajjanakurike Nagaraju Ananda Babu Jayachandra Andrzej Stateczny Swathi Holalu Yogesh Raviprakash Madenur Lingaraju Balaji Prabhu Baluvaneralu Veeranna |
| author_sort | Vinaykumar Vajjanakurike Nagaraju |
| collection | DOAJ |
| description | Land use land cover (LULC) classification using satellite images is crucial for land-use inventories and environment modeling. The LULC classification is a difficult task because of the high dimensional feature space, which affects the classification accuracy. This article proposes a dual strategy-based bald eagle search (DSBES) algorithm and stacked long short-term memory (LSTM) with residual connection for LULC classification. The dual strategy includes adaptive inertia weight and phasor operator strategy to select relevant features from the feature subset. The stacked LSTM contains multiple layers stacked on top of each other to capture high-level temporal data. By integrating residual connection with stacked LSTM, gradient flow is enabled directly among long sequences, reducing the vanishing gradient issue and fastening the convergence rate. The DSBES and stacked LSTM with residual connection performance are examined in terms of metrics of accuracy, precision, sensitivity, specificity, f1-score, and computational time. The DSBES and stacked LSTM with residual connection achieve higher accuracy values of 99.71%, 98.66%, 97.59%, and 99.24% for UCM, AID, NWPU, and EuroSAT datasets, respectively, as compared to VGG19 and optimal guidance whale optimization algorithm–bidirectional long short-term memory. |
| format | Article |
| id | doaj-art-51716f60643842a3afd1cc0dcbe7eb21 |
| institution | Kabale University |
| issn | 1939-1404 2151-1535 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj-art-51716f60643842a3afd1cc0dcbe7eb212025-01-30T00:00:13ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01184188419810.1109/JSTARS.2024.352219710815620Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover ClassificationVinaykumar Vajjanakurike Nagaraju0Ananda Babu Jayachandra1Andrzej Stateczny2https://orcid.org/0000-0002-4671-6827Swathi Holalu Yogesh3Raviprakash Madenur Lingaraju4Balaji Prabhu Baluvaneralu Veeranna5Department of Information Science and Engineering, Malnad College of Engineering, Hassan, IndiaDepartment of Information Science and Engineering, Malnad College of Engineering, Hassan, IndiaDepartment of Navigation, Gdynia Maritime University, Gdynia, PolandDepartment of Computer Science and Engineering (Artificial Intelligence and Machine Learning), Malnad College of Engineering, Hassan, IndiaDepartment of Artificial Intelligence and Machine Learning, Kalpataru Institute of Technology, Tiptur, IndiaDepartment of Computer Science and Engineering (Artificial Intelligence and Machine Learning), Malnad College of Engineering, Hassan, IndiaLand use land cover (LULC) classification using satellite images is crucial for land-use inventories and environment modeling. The LULC classification is a difficult task because of the high dimensional feature space, which affects the classification accuracy. This article proposes a dual strategy-based bald eagle search (DSBES) algorithm and stacked long short-term memory (LSTM) with residual connection for LULC classification. The dual strategy includes adaptive inertia weight and phasor operator strategy to select relevant features from the feature subset. The stacked LSTM contains multiple layers stacked on top of each other to capture high-level temporal data. By integrating residual connection with stacked LSTM, gradient flow is enabled directly among long sequences, reducing the vanishing gradient issue and fastening the convergence rate. The DSBES and stacked LSTM with residual connection performance are examined in terms of metrics of accuracy, precision, sensitivity, specificity, f1-score, and computational time. The DSBES and stacked LSTM with residual connection achieve higher accuracy values of 99.71%, 98.66%, 97.59%, and 99.24% for UCM, AID, NWPU, and EuroSAT datasets, respectively, as compared to VGG19 and optimal guidance whale optimization algorithm–bidirectional long short-term memory.https://ieeexplore.ieee.org/document/10815620/Dual strategy-based bald eagle search (DSBES)land use land cover (LULC)phasor operatorregional variabilityresidual connectionstacked long short-term memory (LSTM) |
| spellingShingle | Vinaykumar Vajjanakurike Nagaraju Ananda Babu Jayachandra Andrzej Stateczny Swathi Holalu Yogesh Raviprakash Madenur Lingaraju Balaji Prabhu Baluvaneralu Veeranna Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Dual strategy-based bald eagle search (DSBES) land use land cover (LULC) phasor operator regional variability residual connection stacked long short-term memory (LSTM) |
| title | Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification |
| title_full | Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification |
| title_fullStr | Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification |
| title_full_unstemmed | Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification |
| title_short | Dual Strategy Based Improved Swarm Intelligence and Stacked LSTM With Residual Connection for Land Use Land Cover Classification |
| title_sort | dual strategy based improved swarm intelligence and stacked lstm with residual connection for land use land cover classification |
| topic | Dual strategy-based bald eagle search (DSBES) land use land cover (LULC) phasor operator regional variability residual connection stacked long short-term memory (LSTM) |
| url | https://ieeexplore.ieee.org/document/10815620/ |
| work_keys_str_mv | AT vinaykumarvajjanakurikenagaraju dualstrategybasedimprovedswarmintelligenceandstackedlstmwithresidualconnectionforlanduselandcoverclassification AT anandababujayachandra dualstrategybasedimprovedswarmintelligenceandstackedlstmwithresidualconnectionforlanduselandcoverclassification AT andrzejstateczny dualstrategybasedimprovedswarmintelligenceandstackedlstmwithresidualconnectionforlanduselandcoverclassification AT swathiholaluyogesh dualstrategybasedimprovedswarmintelligenceandstackedlstmwithresidualconnectionforlanduselandcoverclassification AT raviprakashmadenurlingaraju dualstrategybasedimprovedswarmintelligenceandstackedlstmwithresidualconnectionforlanduselandcoverclassification AT balajiprabhubaluvaneraluveeranna dualstrategybasedimprovedswarmintelligenceandstackedlstmwithresidualconnectionforlanduselandcoverclassification |