摘要
网箱养殖是海水养殖中最重要的类型之一,各类网箱在遥感影像中形状不一,且背景复杂,以往的网箱提取方法,未能完全模拟人类的视觉行为,以及高效利用光谱信息。针对上述问题,提出深度多循环注意力光谱的U-Net网络模型(Spectral Loopy Attention U-Net,SLA-UNet)进行网箱养殖信息提取,使用基于最优尺度寻优(Estimation of Scale Parameter,ESP)的随机森林(Random Forest,RF)算法,去除波段运算后的冗余光谱信息,并添加类似人眼的注意力行为机制,深化影响网箱信息提取的重要特征通道,同时进行边缘补齐补充损失信息,实现了网箱养殖信息的高精度提取。选取广东省湛江市和海南省临高县作为研究区域,与Canny算子、Otsu算法、PCA_Kmeans算法、基于ESP的RF算法、U-Net模型提取结果进行对比,所提SLA-UNet模型近岸网箱的提取精度为98.3%,深海网箱提取精度平均值为98.9%,验证了SLA-UNet模型在网箱养殖识别中的有效性。
Cage aquaculture is one of the most important types of marine aquaculture.Different types of cages have varying shapes in remote sensing images,and the background is complex.Previous methods for cage extraction have not been able to fully simulate human visual behavior and efficiently utilize spectral information.To address these issues,we propose a Spectral Loopy Attention U-Net(SLA-UNet) network model for cage aquaculture information extraction.The model utilizes the Random Forest(RF) algorithm based on the Estimation of Scale Parameter(ESP) to remove redundant spectral information after band operations.It also incorporates a human-like attention mechanism to enhance the important feature channels that affect cage information extraction.Additionally,edge completion is performed to supplement the loss information,achieving high-precision extraction of cage aquaculture information.We selected Zhanjiang City,Guangdong Province and Lingao County,as the study areas.Comparisons were made with the extraction results of the Canny algorithm,Otsu algorithm,PCA_Kmeans algorithm,RF algorithm based on ESP,and the U-Net model.The extraction accuracy of the SLA-UNet model for nearshore cages is 98.3%,and the average extraction accuracy for deep-sea cages is 98.9%,validating the effectiveness of the SLA-UNet model in cage aquaculture recognition.
作者
柯丽娜
由金浩
范剑超
Ke Li’na;You Jinhao;Fan Jianchao(School of Geographical,Liaoning Normal University,Dalian 116029,China;School of Control Science and Engineering,Dalian University of Technology,Dalian 116023,China)
出处
《海洋学报》
CAS
CSCD
北大核心
2024年第5期93-102,共10页
基金
国家自然科学基金项目(42076184,41706195)
国家高分重大科研专项项目(41-Y30F07-9001-20/22)
国家重点研发计划项目(2021YFC2801000)。
关键词
网箱养殖
U-Net模型
多循环注意力机制
深度特征
高效光谱特征
cage aquaculture
U-Net model
Loopy Attention mechanism
deep features
highly efficient spectral characteristics
