Deep ensemble model-based moving object detection and classification using SAR images
In recent decades, image processing and computer vision models have played a vital role in moving object detection on the synthetic aperture radar (SAR) images. Capturing of moving objects in the SAR images is a difficult task. In this study, a new automated model for detecting moving objects is pro...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2024-01-01
|
| Series: | Frontiers in Earth Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/feart.2023.1288003/full |
| _version_ | 1797352586094837760 |
|---|---|
| author | Ramya Paramasivam Prashanth Kumar Wen-Cheng Lai Parameshachari Bidare Divakarachari |
| author_facet | Ramya Paramasivam Prashanth Kumar Wen-Cheng Lai Parameshachari Bidare Divakarachari |
| author_sort | Ramya Paramasivam |
| collection | DOAJ |
| description | In recent decades, image processing and computer vision models have played a vital role in moving object detection on the synthetic aperture radar (SAR) images. Capturing of moving objects in the SAR images is a difficult task. In this study, a new automated model for detecting moving objects is proposed using SAR images. The proposed model has four main steps, namely, preprocessing, segmentation, feature extraction, and classification. Initially, the input SAR image is pre-processed using a histogram equalization technique. Then, the weighted Otsu-based segmentation algorithm is applied for segmenting the object regions from the pre-processed images. When using the weighted Otsu, the segmented grayscale images are not only clear but also retain the detailed features of grayscale images. Next, feature extraction is carried out by gray-level co-occurrence matrix (GLCM), median binary patterns (MBPs), and additive harmonic mean estimated local Gabor binary pattern (AHME-LGBP). The final step is classification using deep ensemble models, where the objects are classified by employing the ensemble deep learning technique, combining the models like the bidirectional long short-term memory (Bi-LSTM), recurrent neural network (RNN), and improved deep belief network (IDBN), which is trained with the features extracted previously. The combined models increase the accuracy of the results significantly. Furthermore, ensemble modeling reduces the variance and modeling method bias, which decreases the chances of overfitting. Compared to a single contributing model, ensemble models perform better and make better predictions. Additionally, an ensemble lessens the spread or dispersion of the model performance and prediction accuracy. Finally, the performance of the proposed model is related to the conventional models with respect to different measures. In the mean-case scenario, the proposed ensemble model has a minimum error value of 0.032, which is better related to other models. In both median- and best-case scenario studies, the ensemble model has a lower error value of 0.029 and 0.015. |
| first_indexed | 2024-03-08T13:17:46Z |
| format | Article |
| id | doaj.art-cd0efd4521d945d4a5c4c14d1da30ee8 |
| institution | Directory Open Access Journal |
| issn | 2296-6463 |
| language | English |
| last_indexed | 2024-03-08T13:17:46Z |
| publishDate | 2024-01-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Earth Science |
| spelling | doaj.art-cd0efd4521d945d4a5c4c14d1da30ee82024-01-18T04:38:00ZengFrontiers Media S.A.Frontiers in Earth Science2296-64632024-01-011110.3389/feart.2023.12880031288003Deep ensemble model-based moving object detection and classification using SAR imagesRamya Paramasivam0Prashanth Kumar1Wen-Cheng Lai2Parameshachari Bidare Divakarachari3Department of Computer Science and Engineering, Mahendra Engineering College (Autonomous), Mallasamudram, IndiaSchool of Information and Physical Sciences, College of Engineering, Science and Environment, University of Newcastle, Callaghan, NSW, AustraliaDepartment of Electrical Engineering, Ming Chi University of Technology, New Taipei City, TaiwanDepartment of Electronics and Communication Engineering, Nitte Meenakshi Institute of Technology, Bengaluru, IndiaIn recent decades, image processing and computer vision models have played a vital role in moving object detection on the synthetic aperture radar (SAR) images. Capturing of moving objects in the SAR images is a difficult task. In this study, a new automated model for detecting moving objects is proposed using SAR images. The proposed model has four main steps, namely, preprocessing, segmentation, feature extraction, and classification. Initially, the input SAR image is pre-processed using a histogram equalization technique. Then, the weighted Otsu-based segmentation algorithm is applied for segmenting the object regions from the pre-processed images. When using the weighted Otsu, the segmented grayscale images are not only clear but also retain the detailed features of grayscale images. Next, feature extraction is carried out by gray-level co-occurrence matrix (GLCM), median binary patterns (MBPs), and additive harmonic mean estimated local Gabor binary pattern (AHME-LGBP). The final step is classification using deep ensemble models, where the objects are classified by employing the ensemble deep learning technique, combining the models like the bidirectional long short-term memory (Bi-LSTM), recurrent neural network (RNN), and improved deep belief network (IDBN), which is trained with the features extracted previously. The combined models increase the accuracy of the results significantly. Furthermore, ensemble modeling reduces the variance and modeling method bias, which decreases the chances of overfitting. Compared to a single contributing model, ensemble models perform better and make better predictions. Additionally, an ensemble lessens the spread or dispersion of the model performance and prediction accuracy. Finally, the performance of the proposed model is related to the conventional models with respect to different measures. In the mean-case scenario, the proposed ensemble model has a minimum error value of 0.032, which is better related to other models. In both median- and best-case scenario studies, the ensemble model has a lower error value of 0.029 and 0.015.https://www.frontiersin.org/articles/10.3389/feart.2023.1288003/fullbidirectional long short-term memoryimproved deep belief networkmoving object detectionrecurrent neural networksynthetic aperture radar images |
| spellingShingle | Ramya Paramasivam Prashanth Kumar Wen-Cheng Lai Parameshachari Bidare Divakarachari Deep ensemble model-based moving object detection and classification using SAR images Frontiers in Earth Science bidirectional long short-term memory improved deep belief network moving object detection recurrent neural network synthetic aperture radar images |
| title | Deep ensemble model-based moving object detection and classification using SAR images |
| title_full | Deep ensemble model-based moving object detection and classification using SAR images |
| title_fullStr | Deep ensemble model-based moving object detection and classification using SAR images |
| title_full_unstemmed | Deep ensemble model-based moving object detection and classification using SAR images |
| title_short | Deep ensemble model-based moving object detection and classification using SAR images |
| title_sort | deep ensemble model based moving object detection and classification using sar images |
| topic | bidirectional long short-term memory improved deep belief network moving object detection recurrent neural network synthetic aperture radar images |
| url | https://www.frontiersin.org/articles/10.3389/feart.2023.1288003/full |
| work_keys_str_mv | AT ramyaparamasivam deepensemblemodelbasedmovingobjectdetectionandclassificationusingsarimages AT prashanthkumar deepensemblemodelbasedmovingobjectdetectionandclassificationusingsarimages AT wenchenglai deepensemblemodelbasedmovingobjectdetectionandclassificationusingsarimages AT parameshacharibidaredivakarachari deepensemblemodelbasedmovingobjectdetectionandclassificationusingsarimages |