A small object detection model in aerial images based on CPDD-YOLOv8
Abstract Aerial images can cover a wide area and capture rich scene information. These images are often taken from a high altitude and contain many small objects. It is difficult to detect small objects accurately because their features are not obvious and are susceptible to background interference....
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-84938-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559698477678592 |
|---|---|
| author | Jingyang Wang Jiayao Gao Bo Zhang |
| author_facet | Jingyang Wang Jiayao Gao Bo Zhang |
| author_sort | Jingyang Wang |
| collection | DOAJ |
| description | Abstract Aerial images can cover a wide area and capture rich scene information. These images are often taken from a high altitude and contain many small objects. It is difficult to detect small objects accurately because their features are not obvious and are susceptible to background interference. The CPDD-YOLOv8 is proposed to improve the performance of small object detection. Firstly, we propose the C2fGAM structure, which integrates the Global Attention Mechanism (GAM) into the C2f structure of the backbone so that the model can better understand the overall semantics of the images. Secondly, a detection layer named P2 is added to extract the shallow features. Thirdly, a new DSC2f structure is proposed, which uses Dynamic Snake Convolution (DSConv) to take the place of the first standard Conv of Bottleneck in the C2f structure, so that the model can adapt to different inputs more effectively. Finally, the Dynamic Head (DyHead), which integrates multiple attention mechanisms, is used in the head to assign different weights to different feature layers. To prove the effectiveness of the CPDD-YOLOv8, we carry out ablation and comparison experiments on the VisDrone2019 dataset. Ablation experiments show that all the improved and added modules in CPDD-YOLOv8 are effective. Comparative experiments suggest that the mAP of CPDD-YOLOv8 is higher than the other seven comparison models. The mAP@0.5 of this model reaches 41%, which is 6.9% higher than that of YOLOv8. The CPDD-YOLOv8’s small object detection rate is improved by 13.1%. The generalizability of the CPDD-YOLOv8 model is verified on the WiderPerson, VOC_MASK and SHWD datasets. |
| format | Article |
| id | doaj-art-de34628601ca4f0db5f732e6aa339af1 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-de34628601ca4f0db5f732e6aa339af12025-01-05T12:17:45ZengNature PortfolioScientific Reports2045-23222025-01-0115111610.1038/s41598-024-84938-4A small object detection model in aerial images based on CPDD-YOLOv8Jingyang Wang0Jiayao Gao1Bo Zhang2School of Information Science and Engineering, Hebei University of Science and TechnologySchool of Information Science and Engineering, Hebei University of Science and TechnologySchool of Cyberspace Security, Hebei University of Engineering ScienceAbstract Aerial images can cover a wide area and capture rich scene information. These images are often taken from a high altitude and contain many small objects. It is difficult to detect small objects accurately because their features are not obvious and are susceptible to background interference. The CPDD-YOLOv8 is proposed to improve the performance of small object detection. Firstly, we propose the C2fGAM structure, which integrates the Global Attention Mechanism (GAM) into the C2f structure of the backbone so that the model can better understand the overall semantics of the images. Secondly, a detection layer named P2 is added to extract the shallow features. Thirdly, a new DSC2f structure is proposed, which uses Dynamic Snake Convolution (DSConv) to take the place of the first standard Conv of Bottleneck in the C2f structure, so that the model can adapt to different inputs more effectively. Finally, the Dynamic Head (DyHead), which integrates multiple attention mechanisms, is used in the head to assign different weights to different feature layers. To prove the effectiveness of the CPDD-YOLOv8, we carry out ablation and comparison experiments on the VisDrone2019 dataset. Ablation experiments show that all the improved and added modules in CPDD-YOLOv8 are effective. Comparative experiments suggest that the mAP of CPDD-YOLOv8 is higher than the other seven comparison models. The mAP@0.5 of this model reaches 41%, which is 6.9% higher than that of YOLOv8. The CPDD-YOLOv8’s small object detection rate is improved by 13.1%. The generalizability of the CPDD-YOLOv8 model is verified on the WiderPerson, VOC_MASK and SHWD datasets.https://doi.org/10.1038/s41598-024-84938-4YOLOv8C2fGAMP2; DSC2fSmall object detection |
| spellingShingle | Jingyang Wang Jiayao Gao Bo Zhang A small object detection model in aerial images based on CPDD-YOLOv8 Scientific Reports YOLOv8 C2fGAM P2; DSC2f Small object detection |
| title | A small object detection model in aerial images based on CPDD-YOLOv8 |
| title_full | A small object detection model in aerial images based on CPDD-YOLOv8 |
| title_fullStr | A small object detection model in aerial images based on CPDD-YOLOv8 |
| title_full_unstemmed | A small object detection model in aerial images based on CPDD-YOLOv8 |
| title_short | A small object detection model in aerial images based on CPDD-YOLOv8 |
| title_sort | small object detection model in aerial images based on cpdd yolov8 |
| topic | YOLOv8 C2fGAM P2; DSC2f Small object detection |
| url | https://doi.org/10.1038/s41598-024-84938-4 |
| work_keys_str_mv | AT jingyangwang asmallobjectdetectionmodelinaerialimagesbasedoncpddyolov8 AT jiayaogao asmallobjectdetectionmodelinaerialimagesbasedoncpddyolov8 AT bozhang asmallobjectdetectionmodelinaerialimagesbasedoncpddyolov8 AT jingyangwang smallobjectdetectionmodelinaerialimagesbasedoncpddyolov8 AT jiayaogao smallobjectdetectionmodelinaerialimagesbasedoncpddyolov8 AT bozhang smallobjectdetectionmodelinaerialimagesbasedoncpddyolov8 |