耕地高分辨率影像多尺度场景分类及非农化应用  

Multi-scale Scene Classification and Non-agricultural Application of Cultivated Land High-resolution Image

作　　者：陈巍 李浩[1] 张启华 何燕兰 王胜利 CHEN Wei;LI Hao;ZHANG Qihua;HE Yanlan;WANG Shengli(School of Earth Science and Engineering,Hohai University,Nanjing 211100,China;Jiangsu Provincial Geological Surveying and Mapping Brigade,Nanjing 211102,China;School of Environment and Spatial Informatics,China University of Mining and Technology,Xuzhou 221116,China)

机构地区：[1]河海大学地球科学与工程学院,江苏南京211100 [2]江苏省地质测绘大队,江苏南京211102 [3]中国矿业大学环境与测绘学院,江苏徐州221116

出　　处：《遥感技术与应用》2025年第1期25-37,共13页Remote Sensing Technology and Application

基　　金：江苏省地质局科研项目(2022KY15)。

摘　　要：快速、准确地获取耕地变化信息对粮食安全管理具有重要意义。针对遥感语义分割方法在大尺度范围高分辨率影像耕地非农化检测中因模型适用性不足导致错漏检较多的问题,提出一种以场景分类网络Xception为基础的耕地高分辨率影像多尺度场景分类方法--Multiscale Scene Classification-Xception(MSC-Xception)。该方法对耕地场景分类性能突出的轻量级场景分类网络Xception的输出层嵌入卷积注意力模块CBAM以增强模型在通道及空间特征上的提取能力,同时对单一尺度场景级分类在大尺度耕地提取中存在的混合场景分离度低和细节粗糙问题,先引入一种多尺度耕地场景特征融合的方法提高混合场景的分离度,再通过多尺度分割矢量的边界约束实现对场景级分类的边界精细化。相较于典型的Unet、PSPNet和DeeplabV3+语义分割方法,该方法能较好地减少大图斑漏检现象,在栖霞区2018年4月份GF-2影像的耕地提取实验中召回率和F1分数至少分别提高了15.1个百分点和8.8个百分点,在2018年至2022年栖霞区耕地非农化检测中,可疑图斑的查全率至少提高了7.16个百分点。Rapid and accurate acquisition of cultivated land change information is of great significance to food se⁃curity management.This paper aims at the problem that remote sensing semantic segmentation method has many errors and omissions due to insufficient model applicability in large-scale and high-resolution image culti⁃vated land non-agricultural detection.Multiscale Scene Classification-Xception(MSC-Xception),a multiscale scene classification method for high-resolution cultivated land images based on Xception,is proposed.The convolutional attention module CBAM is embedded into the output layer of the lightweight scene classifica⁃tion network Xception,which has outstanding performance in cultivated land scene classification,to enhance the model's ability to extract channel and spatial features.At the same time,the problem of low separation de⁃gree and rough details of mixed scenes existing in the single-scale scene-level classification in large-scale culti⁃vated land extraction is also overcome.Firstly,a feature fusion method of multi-scale cultivated land scene is in⁃troduced to improve the separation degree of mixed scene,and then the boundary constraint of multi-scale seg⁃mentation vector is used to achieve the boundary refinement of scene-level classification.Compared with the typical Unet,PSPNet and DeeplabV3+semantic segmentation methods,this method can better reduce the missed detection of large map spots,and the recall rate and F1 score in the cultivated land extraction experiment of GF-2 images in Qixia District in April 2018 increased by at least 15.1 percentage points and 8.8 percentage points respectively.In the non-agricultural detection of cultivated land in Qixia District from 2018 to 2022,the recall rate of suspicious spots increased by at least 7.16 percentage points.

关 键 词：非农化 耕地 多尺度 场景分类 高分辨率 遥感 Xception 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] S127[自动化与计算机技术—控制科学与工程]