Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images
To overcome the limitations of traditional image alignment methods in capturing deep semantic features, a deep feature information image alignment network (DFA-Net) is proposed. This network aims to enhance image alignment performance through multi-level feature learning. DFA-Net is based on the dee...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/15/4628 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849406404744445952 |
|---|---|
| author | Lin Zhu Yuxing Mao Jianyu Pan |
| author_facet | Lin Zhu Yuxing Mao Jianyu Pan |
| author_sort | Lin Zhu |
| collection | DOAJ |
| description | To overcome the limitations of traditional image alignment methods in capturing deep semantic features, a deep feature information image alignment network (DFA-Net) is proposed. This network aims to enhance image alignment performance through multi-level feature learning. DFA-Net is based on the deep residual architecture and introduces spatial pyramid pooling to achieve cross-scalar feature fusion, effectively enhancing the feature’s adaptability to scale. A feature enhancement module based on the self-attention mechanism is designed, with key features that exhibit geometric invariance and high discriminative power, achieved through a dynamic weight allocation strategy. This improves the network’s robustness to multimodal image deformation. Experiments on two public datasets, MSRS and RoadScene, show that the method performs well in terms of alignment accuracy, with the RMSE metrics being reduced by 0.661 and 0.473, and the SSIM, MI, and NCC improved by 0.155, 0.163, and 0.211; and 0.108, 0.226, and 0.114, respectively, compared with the benchmark model. The visualization results validate the significant improvement in the features’ visual quality and confirm the method’s advantages in terms of stability and discriminative properties of deep feature extraction. |
| format | Article |
| id | doaj-art-be4cada7b63247ee8008cc1be89ed20c |
| institution | Kabale University |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-be4cada7b63247ee8008cc1be89ed20c2025-08-20T03:36:23ZengMDPI AGSensors1424-82202025-07-012515462810.3390/s25154628Image Alignment Based on Deep Learning to Extract Deep Feature Information from ImagesLin Zhu0Yuxing Mao1Jianyu Pan2State Key Laboratory of Power Transmission Equipment Technology, School of Electrical Engineering, Chongqing University, Chongqing 400044, ChinaState Key Laboratory of Power Transmission Equipment Technology, School of Electrical Engineering, Chongqing University, Chongqing 400044, ChinaState Key Laboratory of Power Transmission Equipment Technology, School of Electrical Engineering, Chongqing University, Chongqing 400044, ChinaTo overcome the limitations of traditional image alignment methods in capturing deep semantic features, a deep feature information image alignment network (DFA-Net) is proposed. This network aims to enhance image alignment performance through multi-level feature learning. DFA-Net is based on the deep residual architecture and introduces spatial pyramid pooling to achieve cross-scalar feature fusion, effectively enhancing the feature’s adaptability to scale. A feature enhancement module based on the self-attention mechanism is designed, with key features that exhibit geometric invariance and high discriminative power, achieved through a dynamic weight allocation strategy. This improves the network’s robustness to multimodal image deformation. Experiments on two public datasets, MSRS and RoadScene, show that the method performs well in terms of alignment accuracy, with the RMSE metrics being reduced by 0.661 and 0.473, and the SSIM, MI, and NCC improved by 0.155, 0.163, and 0.211; and 0.108, 0.226, and 0.114, respectively, compared with the benchmark model. The visualization results validate the significant improvement in the features’ visual quality and confirm the method’s advantages in terms of stability and discriminative properties of deep feature extraction.https://www.mdpi.com/1424-8220/25/15/4628image alignmentdeep learningfeature extractioninfrared and visible images |
| spellingShingle | Lin Zhu Yuxing Mao Jianyu Pan Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images Sensors image alignment deep learning feature extraction infrared and visible images |
| title | Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images |
| title_full | Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images |
| title_fullStr | Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images |
| title_full_unstemmed | Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images |
| title_short | Image Alignment Based on Deep Learning to Extract Deep Feature Information from Images |
| title_sort | image alignment based on deep learning to extract deep feature information from images |
| topic | image alignment deep learning feature extraction infrared and visible images |
| url | https://www.mdpi.com/1424-8220/25/15/4628 |
| work_keys_str_mv | AT linzhu imagealignmentbasedondeeplearningtoextractdeepfeatureinformationfromimages AT yuxingmao imagealignmentbasedondeeplearningtoextractdeepfeatureinformationfromimages AT jianyupan imagealignmentbasedondeeplearningtoextractdeepfeatureinformationfromimages |