遥感影像耕地提取的研究进展与展望
Research Progress and Prospect of Cultivated Land Extraction from Remote Sensing Images
巫志雄 1李乔宇 2王宗良 3曾世伟3
作者信息
- 1. 聊城大学物理科学与信息工程学院,山东 聊城 252000;山东省农业科学院农业信息与经济研究所,山东 济南 250100
- 2. 山东省农业科学院农业信息与经济研究所,山东 济南 250100
- 3. 聊城大学物理科学与信息工程学院,山东 聊城 252000
- 折叠
摘要
获取实时、精准的耕地分布信息是现代土地资源管理和农业高质量发展中至关重要的一项任务.随着卫星技术的迅猛发展,遥感监测逐渐成为当前耕地信息提取的重要手段,同时,深度学习技术迅速崛起,并逐渐成为遥感影像耕地提取的关键技术.本文整理了国内外近期耕地提取的相关研究成果,阐述了传统提取算法的不足及高分辨率遥感图像对耕地提取的积极意义、耕地提取的基本流程、耕地提取算法发展的主要过程和研究策略,归纳了耕地提取算法的主要优化方法以及多任务网络模型的应用,最后结合现有深度学习算法存在的不足对未来耕地提取技术发展趋势进行了展望.
Abstract
Obtaining real-time and accurate information on cultivated land distribution is a critical task in modern land resource management and high-quality agricultural development.With the rapid development of satellite technology,remote sensing monitoring has gradually become an important method for extracting culti-vated land information.At the same time,deep learning technology has risen rapidly and is increasingly be-coming a pivotal technique for cultivated land extraction from remote sensing imagers.In this paper,recent re-search findings on cultivated land extraction both domestically and internationally were summarized,the limita-tions of traditional extraction algorithms,and the positive significance of high-resolution remote sensing images to cultivated land extraction were elaborated.The fundamental processes of cultivated land extraction were out-lined,and the primary development stages and research strategies of cultivated land extraction algorithms were reviewed,and the main optimization methods for such algorithms,and the application of multi-task network models were summarized.Finally,the limitations of current deep learning algorithms were discussed and future trends in the development of cultivated land extraction technologies were anticipated.
关键词
耕地提取/深度学习/语义分割/高分辨率影像Key words
Cultivated land extraction/Deep learning/Semantic segmentation/High resolution image引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据