摘要
基于多源低空遥感数据获取、知识样本库、深度学习变化感知、多源数据融合分析,提出以专题“样本字典”和自注意力机制变化感知为核心的快速监测服务技术体系,旨在解决国家重要战略发展地区违法用地监管高效、精准、精细的管控要求,有效保障土地执法效率及准确率。
Based on multi-source low altitude remote sensing data acquisition,knowledge sample base,deep learning change perception,multi-source data fusion analysis,we proposed a rapid monitoring service technology system which taking the special“Sample Dictionary”and self-attention mechanism change perception as the core,so as to meet the requirements of efficient,accurate and fine management and control of illegal land use supervision in important national strategic development areas,and effectively protect the efficiency and accuracy of land use law enforcement.
出处
《地理空间信息》
2021年第6期128-131,I0003,共5页
Geospatial Information
关键词
低空遥感
样本字典
自注意力机制模型
深度学习
智能感知
low altitude remote sensing
“Sample Dictionary”
self-attention mechanism model
deep learning
intelligent perception
