Maqola

Lightweight GAN for Restoring Blurred Images to Enhance Citrus Detection

Yuyu HuangKey Laboratory of Intelligent Agricultural Equipment in Hilly and Mountain Areas, College of Engineering and Technology, Southwest University, Chongqing 400715, ChinaHui LiKey Laboratory of Intelligent Agricultural Equipment in Hilly and Mountain Areas, College of Engineering and Technology, Southwest University, Chongqing 400715, ChinaYuheng YangKey Laboratory of Agricultural Biosafety and Green Production of Upper Yangtze River, College of Plant Protection, Southwest University, Ministry of Education, Chongqing 400715, ChinaChengsong LiKey Laboratory of Intelligent Agricultural Equipment in Hilly and Mountain Areas, College of Engineering and Technology, Southwest University, Chongqing 400715, ChinaLihong WangKey Laboratory of Intelligent Agricultural Equipment in Hilly and Mountain Areas, College of Engineering and Technology, Southwest University, Chongqing 400715, ChinaPei WangKey Laboratory of Intelligent Agricultural Equipment in Hilly and Mountain Areas, College of Engineering and Technology, Southwest University, Chongqing 400715, China

2025en

ABI

Annotatsiya

Image blur is a major factor that degrades object detection in agricultural applications, particularly in orchards where crop occlusion, leaf movement, and camera shake frequently reduce image quality. This study proposed a lightweight generative adversarial network, AGG-DeblurGAN, to address non-uniform motion blur in citrus tree images. The model integrates the GhostNet backbone, attention-enhanced Ghost modules, and a Gated Half Instance Normalization Module. A blur detection mechanism enabled dynamic routing, reducing computation on sharp images. Experiments on a citrus dataset showed that AGG-DeblurGAN maintained restoration quality while improving efficiency. For object detection, restored citrus images achieved an 86.4% improvement in [email protected]:0.95, a 76.9% gain in recall, and a 40.1% increase in F1 score compared to blurred images, while the false negative rate dropped by 63.9%. These results indicate that AGG-DeblurGAN can serve as a reference for improving image preprocessing and detection performance in agricultural vision systems.

Hali tarjima qilinmagan

Identifikatorlar

DOI: 10.3390/plants14193085

Iqtiboslar va manbalar

2 ta iqtibos0 ta foydalanilgan manba

Koʻrsatkichlar — AkademScholar