基于深度强化学习的图像超分辨率增强系统设计  

Design of Image Super-resolution Enhancement System Based on Deep Reinforcement Learning

作　　者：董立媛 DONG Liyuan(Dezhou Vocational and Technical College,Dezhou Shangdong 253034)

机构地区：[1]德州职业技术学院,山东德州253034

出　　处：《软件》2025年第2期52-54,共3页Software

摘　　要：本文提出了一种基于深度强化学习的图像超分辨率增强系统,结合深度卷积神经网络(DCNN)和强化学习技术,提升了低分辨率图像的恢复质量与处理效率。系统通过图像预处理、特征提取和强化学习优化三个模块,显著改善了图像细节恢复,且在不同数据集上表现优异。测试结果证明,该系统适用于医学影像、遥感监测和安防等领域。This paper presents an image super-resolution enhancement system based on deep reinforcement learning,which combines deep convolutional neural networks(DCNN)and reinforcement learning to improve the restoration quality and processing efficiency of low-resolution images.The system features three modules:image preprocessing,feature extraction,and reinforcement learning optimization,significantly enhancing image detail recovery and performing well across different datasets.Test results show that the system is suitable for applications in medical imaging,remote sensing,and security.

关 键 词：深度强化学习 图像超分辨率增强系统 深度卷积神经网络(DCNN) 强化学习 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]