基于地膜指数与遥感影像特征的农田地膜精准识别OA

To address the current challenges of insufficient feature utilization and limited accuracy in remote sensing identification of agricultural plastic mulch,focusing on Qixian county,Kaifeng city,Henan province as the core study area,firstly,a novel remote sensing index——the plastic film index(PFI)was developed.Simultaneously,multi-dimensional feature sets encompassing spectral data,indices,texture and surface temperature were systematically extracted from Sentinel-2 and Landsat-8 satellite imagery.Building upon this foundation,a parameter-optimized recursive feature elimination(RFE)algorithm was employed for feature selection to eliminate redundant information.Finally,a random forest classifier was utilized to extract the distribution information of agricultural plastic film.The results indicated that the RFE algorithm optimized the initial 65 feature variables to 39,significantly enhancing the model's computational efficiency and robustness.The classification model based on optimized features achieved an overall accuracy of 92%with a Kappa coefficient of 0.89,and both producer and receiver accuracy for plastic film categories exceeded 90%.Compared with other approaches,the model incorporating the PFI index and feature selection demonstrated optimal recognition performance,confirming the effectiveness of the PFI index in enhancing plastic film identification accuracy.Above results not only provided precise spatial information for agricultural producers in residual film recovery and field management,but also offered reliable technical support and data backing for pollution assessment and decision-making by agricultural environmental protection departments.