High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors
Deep-learning-based object detectors have substantially improved state-of-the-art object detection in remote sensing images in terms of precision and degree of automation. Nevertheless, the large variation of the object scales makes it difficult to achieve high-quality detection across multiresoluti...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-04-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/9/2091 |
| _version_ | 1797503082902323200 |
|---|---|
| author | Binglong Wu Yuan Shen Shanxin Guo Jinsong Chen Luyi Sun Hongzhong Li Yong Ao |
| author_facet | Binglong Wu Yuan Shen Shanxin Guo Jinsong Chen Luyi Sun Hongzhong Li Yong Ao |
| author_sort | Binglong Wu |
| collection | DOAJ |
| description | Deep-learning-based object detectors have substantially improved state-of-the-art object detection in remote sensing images in terms of precision and degree of automation. Nevertheless, the large variation of the object scales makes it difficult to achieve high-quality detection across multiresolution remote sensing images, where the quality is defined by the Intersection over Union (IoU) threshold used in training. In addition, the imbalance between the positive and negative samples across multiresolution images worsens the detection precision. Recently, it was found that a Cascade region-based convolutional neural network (R-CNN) can potentially achieve a higher quality of detection by introducing a cascaded three-stage structure using progressively improved IoU thresholds. However, the performance of Cascade R-CNN degraded when the fourth stage was added. We investigated the cause and found that the mismatch between the ROI features and the classifier could be responsible for the degradation of performance. Herein, we propose a Cascade R-CNN++ structure to address this issue and extend the three-stage architecture to multiple stages for general use. Specifically, for cascaded classification, we propose a new ensemble strategy for the classifier and region of interest (RoI) features to improve classification accuracy at inference. In localization, we modified the loss function of the bounding box regressor to obtain higher sensitivity around zero. Experiments on the DOTA dataset demonstrated that Cascade R-CNN++ outperforms Cascade R-CNN in terms of precision and detection quality. We conducted further analysis on multiresolution remote sensing images to verify model transferability across different object scales. |
| first_indexed | 2024-03-10T03:45:26Z |
| format | Article |
| id | doaj.art-f2dd7a03ec354c3cbbe10f4006b8c0c7 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T03:45:26Z |
| publishDate | 2022-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-f2dd7a03ec354c3cbbe10f4006b8c0c72023-11-23T09:10:26ZengMDPI AGRemote Sensing2072-42922022-04-01149209110.3390/rs14092091High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage DetectorsBinglong Wu0Yuan Shen1Shanxin Guo2Jinsong Chen3Luyi Sun4Hongzhong Li5Yong Ao6Center for Geo-Spatial Information, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaCenter for Geo-Spatial Information, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaCenter for Geo-Spatial Information, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaCenter for Geo-Spatial Information, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaCenter for Geo-Spatial Information, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaCenter for Geo-Spatial Information, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaSchool of Earth Science and Resources, Chang’an University, 126 Yanta Road, Xi’an 710054, ChinaDeep-learning-based object detectors have substantially improved state-of-the-art object detection in remote sensing images in terms of precision and degree of automation. Nevertheless, the large variation of the object scales makes it difficult to achieve high-quality detection across multiresolution remote sensing images, where the quality is defined by the Intersection over Union (IoU) threshold used in training. In addition, the imbalance between the positive and negative samples across multiresolution images worsens the detection precision. Recently, it was found that a Cascade region-based convolutional neural network (R-CNN) can potentially achieve a higher quality of detection by introducing a cascaded three-stage structure using progressively improved IoU thresholds. However, the performance of Cascade R-CNN degraded when the fourth stage was added. We investigated the cause and found that the mismatch between the ROI features and the classifier could be responsible for the degradation of performance. Herein, we propose a Cascade R-CNN++ structure to address this issue and extend the three-stage architecture to multiple stages for general use. Specifically, for cascaded classification, we propose a new ensemble strategy for the classifier and region of interest (RoI) features to improve classification accuracy at inference. In localization, we modified the loss function of the bounding box regressor to obtain higher sensitivity around zero. Experiments on the DOTA dataset demonstrated that Cascade R-CNN++ outperforms Cascade R-CNN in terms of precision and detection quality. We conducted further analysis on multiresolution remote sensing images to verify model transferability across different object scales.https://www.mdpi.com/2072-4292/14/9/2091object detectioncascaded detectorsIntersection over Union (IoU) thresholdclassification ensemblebounding box regressionmultiresolution remote sensing images |
| spellingShingle | Binglong Wu Yuan Shen Shanxin Guo Jinsong Chen Luyi Sun Hongzhong Li Yong Ao High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors Remote Sensing object detection cascaded detectors Intersection over Union (IoU) threshold classification ensemble bounding box regression multiresolution remote sensing images |
| title | High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors |
| title_full | High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors |
| title_fullStr | High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors |
| title_full_unstemmed | High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors |
| title_short | High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors |
| title_sort | high quality object detection for multiresolution remote sensing imagery using cascaded multi stage detectors |
| topic | object detection cascaded detectors Intersection over Union (IoU) threshold classification ensemble bounding box regression multiresolution remote sensing images |
| url | https://www.mdpi.com/2072-4292/14/9/2091 |
| work_keys_str_mv | AT binglongwu highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors AT yuanshen highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors AT shanxinguo highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors AT jinsongchen highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors AT luyisun highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors AT hongzhongli highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors AT yongao highqualityobjectdetectionformultiresolutionremotesensingimageryusingcascadedmultistagedetectors |