首页|Spectral matching based remote sensing identification of two main crop rotation patterns in a large irrigation district
Spectral matching based remote sensing identification of two main crop rotation patterns in a large irrigation district
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
万方数据
维普
The rapid identification of planting patterns for major crops in a large irrigated district has vital importance for irri-gation management,water fee collection,and crop yield estimation.In this study,the OTSU algorithm and Mean-Shift algo-rithm were employed to automatically determine threshold values for mapping two main rotated crop patterns at the pixel scale.A time series analysis was conducted to extract the spatial distribution of rice-wheat and wheat-maize rotations in the Chuan-hang irrigation district from 2016 to 2020.The results demonstrate that both threshold segmentation algorithms are reliable in extracting the spatial distribution of the crops,with an overall accuracy exceeding 80％.Additionally,both Kappa coefficients surpass 0.7,indicating better performance by OTSU method.Over the period from 2016 to 2020,the area occupied by rice-wheat rotation cropping ranged from 12500 to 14400 hm2;whereas wheat-maize rotation cropping exhibited smaller and more variable areas ranging from 19730 to 34070 hm2.These findings highlight how remote sensing-based approaches can provide reliable support for rapidly and accurately identifying the spatial distribution of main crop rotation patterns within a large irri-gation district.
China Institute of Water Resources and Hydropower Research,Beijing 100038,China
Application Center of Remote Sensing Technology,Research Center on Flood & Drought Disaster Prevention and Reduction of the Ministry of Water Resources,Beijing 100038,China
万方数据
维普
