农业信息系统监测中遥感和深度学习技术的应用研究进展OA

In the process of agricultural modernization,the measurement of crop area and estimation of yield play a crucial role in agricultural production management,food security assessment,and the formulation of food policies.However,tradi-tional manual survey methods have problems such as low efficiency,high cost,and lagging data updates,which make it diffi-cult to meet the needs of large-scale,dynamic,and high-precision agricultural monitoring over large areas.In the fields of agricultural remote sensing and deep learning,crop information extraction based on remote sensing technology has been an important research direction in the past decade,and considerable progress has also been made in crop area measurement and yield estimation.This paper systematically reviews the research progress in remote sensing data acquisition and fusion meth-ods,image segmentation and classification methods,and crop area measurement and yield estimation models,and conducts a comparative analysis of the performance of different deep learning models(such as U-Net,SegNet,and DeepLab series,etc.)in crop identification and area estimation.It summarizes the typical applications of remote sensing and deep learning tech-nologies in the field of agricultural information system monitoring at home and abroad,and points out the challenges existing in current research,such as insufficient model generalization ability,inadequate multisource data fusion,and poor model in-terpretability.Combined with the development trend of agricultural digitization,this paper looks forward to the development direction of intelligent remote sensing in future agricultural resource surveys and precise policy implementation to provide inspiration for the current research and practice in the fields related to crop area measurement and yield estimation.