Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images
In recent years, remote sensing images has become one of the most popular directions in image processing. A small feature gap exists between satellite and natural images. Therefore, deep learning algorithms could be applied to recognize remote sensing images. We propose an improved Mask R-CNN model,...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-04-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/8/2618 |
| _version_ | 1827695625756475392 |
|---|---|
| author | Qifan Wu Daqiang Feng Changqing Cao Xiaodong Zeng Zhejun Feng Jin Wu Ziqiang Huang |
| author_facet | Qifan Wu Daqiang Feng Changqing Cao Xiaodong Zeng Zhejun Feng Jin Wu Ziqiang Huang |
| author_sort | Qifan Wu |
| collection | DOAJ |
| description | In recent years, remote sensing images has become one of the most popular directions in image processing. A small feature gap exists between satellite and natural images. Therefore, deep learning algorithms could be applied to recognize remote sensing images. We propose an improved Mask R-CNN model, called SCMask R-CNN, to enhance the detection effect in the high-resolution remote sensing images which contain the dense targets and complex background. Our model can perform object recognition and segmentation in parallel. This model uses a modified SC-conv based on the ResNet101 backbone network to obtain more discriminative feature information and adds a set of dilated convolutions with a specific size to improve the instance segmentation effect. We construct WFA-1400 based on the DOTA dataset because of the shortage of remote sensing mask datasets. We compare the improved algorithm with other state-of-the-art algorithms. The object detection AP<sub>50</sub> and AP increased by 1–2% and 1%, respectively, objectively proving the effectiveness and the feasibility of the improved model. |
| first_indexed | 2024-03-10T12:29:53Z |
| format | Article |
| id | doaj.art-6163308d286246bc9c83d2ebe3e61910 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T12:29:53Z |
| publishDate | 2021-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-6163308d286246bc9c83d2ebe3e619102023-11-21T14:42:18ZengMDPI AGSensors1424-82202021-04-01218261810.3390/s21082618Improved Mask R-CNN for Aircraft Detection in Remote Sensing ImagesQifan Wu0Daqiang Feng1Changqing Cao2Xiaodong Zeng3Zhejun Feng4Jin Wu5Ziqiang Huang6School of Physics and Optoelectronic Engineering, Xidian University, 2 South Taibai Road, Xi’an 710071, ChinaShandong Institute of Space Electronic Technology, Yantai 264670, ChinaSchool of Physics and Optoelectronic Engineering, Xidian University, 2 South Taibai Road, Xi’an 710071, ChinaSchool of Physics and Optoelectronic Engineering, Xidian University, 2 South Taibai Road, Xi’an 710071, ChinaSchool of Physics and Optoelectronic Engineering, Xidian University, 2 South Taibai Road, Xi’an 710071, ChinaSchool of Physics and Optoelectronic Engineering, Xidian University, 2 South Taibai Road, Xi’an 710071, ChinaSchool of Physics and Optoelectronic Engineering, Xidian University, 2 South Taibai Road, Xi’an 710071, ChinaIn recent years, remote sensing images has become one of the most popular directions in image processing. A small feature gap exists between satellite and natural images. Therefore, deep learning algorithms could be applied to recognize remote sensing images. We propose an improved Mask R-CNN model, called SCMask R-CNN, to enhance the detection effect in the high-resolution remote sensing images which contain the dense targets and complex background. Our model can perform object recognition and segmentation in parallel. This model uses a modified SC-conv based on the ResNet101 backbone network to obtain more discriminative feature information and adds a set of dilated convolutions with a specific size to improve the instance segmentation effect. We construct WFA-1400 based on the DOTA dataset because of the shortage of remote sensing mask datasets. We compare the improved algorithm with other state-of-the-art algorithms. The object detection AP<sub>50</sub> and AP increased by 1–2% and 1%, respectively, objectively proving the effectiveness and the feasibility of the improved model.https://www.mdpi.com/1424-8220/21/8/2618Mask R-CNNself-calibrationDOTA datasetaircraftremote sensing image |
| spellingShingle | Qifan Wu Daqiang Feng Changqing Cao Xiaodong Zeng Zhejun Feng Jin Wu Ziqiang Huang Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images Sensors Mask R-CNN self-calibration DOTA dataset aircraft remote sensing image |
| title | Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images |
| title_full | Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images |
| title_fullStr | Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images |
| title_full_unstemmed | Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images |
| title_short | Improved Mask R-CNN for Aircraft Detection in Remote Sensing Images |
| title_sort | improved mask r cnn for aircraft detection in remote sensing images |
| topic | Mask R-CNN self-calibration DOTA dataset aircraft remote sensing image |
| url | https://www.mdpi.com/1424-8220/21/8/2618 |
| work_keys_str_mv | AT qifanwu improvedmaskrcnnforaircraftdetectioninremotesensingimages AT daqiangfeng improvedmaskrcnnforaircraftdetectioninremotesensingimages AT changqingcao improvedmaskrcnnforaircraftdetectioninremotesensingimages AT xiaodongzeng improvedmaskrcnnforaircraftdetectioninremotesensingimages AT zhejunfeng improvedmaskrcnnforaircraftdetectioninremotesensingimages AT jinwu improvedmaskrcnnforaircraftdetectioninremotesensingimages AT ziqianghuang improvedmaskrcnnforaircraftdetectioninremotesensingimages |