.png){kind=icon}
返回
Projectile Explosion Weak Firelight Image Recognition Method Using Multi-Scale Adaptive Enhancement Network
.png){kind=icon}
OA
AI
DOI：10.1109/ACCESS.2024.3505295.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
In order to solve the problem of weak firelight recognition of projectile explosion image obtained by the UAV remote sensing imaging mechanism, this paper proposes a projectile explosion weak firelight recognition method based on multi-scale adaptive enhancement network. The traditional VGG16 backbone network is used as the feature extraction network, and the firelight feature extraction network structure of the projectile explosion image is designed by adding the bidirectional aggregation feature pyramid. Multiple parallel adaptive attention branches are integrated into the structure network, and the contextual enhancement module is used to further optimize the extraction of weak firelight feature detail based on multi-scale projectile explosion image. The Focal-Dice Loss is used as the joint supervision loss function of the network model, which can effectively solve the problem of sample imbalance and improve the recognition accuracy of the network. Based on the UAV remote sensing test platform, this paper collected multi-frame frequency projectile explosion weak firelight images, and verified the weak firelight recognition method. Through comparison and analysis, it is found that the accuracy of the weak firelight recognition method proposed in this paper can reach 95.27%.
Keyword:
Projectiles
Explosions
Feature extraction
Image recognition
Lasers
Target recognition
Adaptive systems
Cameras
Accuracy
Autonomous aerial vehicles
Projectile explosion image
firelight recognition
multi-scale adaptive enhancement network
VGG16
期刊
IF:
3.6
论文数:
9.8W
被引数:
29.4W
机构
暂无机构信息
引用论文
分享收藏SOIF-DN: Preserving Small Object Information Flow With Improved Deep Learning Model.png)
IEEE ACCESSOAAI
IEEE ACCESSIF3.6
OAAI分享收藏Three-dimensional coordinates test method with uncertain projectile proximity explosion position based on dynamic seven photoelectric detection screenDEFENCE TECHNOLOGY
OAAI
DEFENCE TECHNOLOGYIF5.9
OAAI分享收藏分享收藏Classic Spotlight: Electron Bifurcation, a Unifying Concept for Energy Conservation in Anaerobes经典聚焦：电子分岔，厌氧生物能量保存的统一概念OAAI
OAAI分享收藏 没有更多内容
没有更多内容