摘要
目前主流的神经网络在面对复杂多样的地物目标时难以精确区分,同时样本数量少、弱监督条件也容易为神经网络带来大量噪声与错误。为此,在分析遥感影像的地物特点后,提出一种基于权重动态变形的双重网络遥感影像分类方法,通过构架灵活、简易却有效的权重动态变形结构,构建经过改进的分类网络与目标识别网络,形成双网络对照的自我验证,从而提高学习性能、修复误差、增补遗漏、提高分类精度。实验结果表明,所提方法在容易实施的基础上,表现出更强的地物认知能力和更强的噪声抵抗能力,即其能够适应各种遥感影像的分类任务,具有较为广阔的应用潜力。
Currently,popular neural networks not only struggle to accurately recognize various types of surface targets but also tend to introduce significant noise and errors when handling limited samples and weak supervision.Therefore,this study proposes a dualnetwork remote sensing image classification method based on dynamic weight deformation,after analyzing the features of remote sensing images.By constructing a flexible,simple,and effective weight dynamic deformation structure,we establish an improved classification network and target recognition network.This introduces the selfverification ability of dual network comparison,thereby enhancing learning performance,error correction,recognition efficiency,supplementing omissions,and improving classification accuracy.Experimental comparisons show that the proposed method is easy to implement and exhibits stronger cognitive ability and noise resistance.It confirms the adaptability of the proposed method to various remote sensing image classification tasks and its vast application potential.
作者
张庆芳
丛铭
韩玲
席江波
荆青青
崔建军
杨成生
任超峰
顾俊凯
许妙忠
陶翊婷
Zhang Qingfang;Cong Ming;Han Ling;Xi Jiangbo;Jing Qingqing;Cui Jianjun;Yang Chengsheng;Ren Chaofeng;Gu Junkai;Xu Miaozhong;Tao Yiting(College of Geology Engineering and Geomatics,Chang’an University,Xi’an 710054,Shaanxi,Chian;China Aero Geophysical Survey&Remote Sensing Center for Land and Resources,Beijing 100083,China;State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan 430079,Hubei,China)
出处
《激光与光电子学进展》
CSCD
北大核心
2024年第8期275-285,共11页
Laser & Optoelectronics Progress
基金
国家级国家重点研发计划子课题(2021YFC3000404-01)
省部级地调项目独立课题(D20201180)
厅局级项目独立课题(SXLK2021-0225)。
关键词
遥感影像分类
神经网络
权重动态变形
双重神经网络
自我验证
remote sensing image classification
neural network
dynamic weight deformation
dual neural network
self verification
