Abstract:
Timely and accurate spatial distribution information of agricultural plastic greenhouse (APGs) is an important basis for agricultural production. In this study, we used G...Show MoreMetadata
Abstract:
Timely and accurate spatial distribution information of agricultural plastic greenhouse (APGs) is an important basis for agricultural production. In this study, we used Gao-Fen 1 (GF-1) images as data sources to analyze the applicability of spectral characteristics and related indices and texture extraction algorithms under different seasons for the identification of plastic greenhouses in view of the unique spatial distribution details of plastic greenhouses. The results showed that: 1) the accuracies of APGs in different seasons was close, indicating that the method using texture features can be independent of season; 2) among the texture features obtained in different ways, the mean value of gray level co-occurrence matrix (GLCM) had the best recognition effect on APGs. And the overall accuracy of APGs in different seasons obtained based on the optimal features was above 80%, and the highest OA reached 87.01% in site C. This method is of great significance for the feature selection and accurate mapping of the spatial distribution of APGs.
Date of Conference: 16-21 July 2023
Date Added to IEEE Xplore: 20 October 2023
ISBN Information: