Cotton is a globally strategic crop that plays a crucial role in sustaining national economies and livelihoods. To address the challenges of accurate cotton field extraction in the complex planting environments of Xinjiang’s Alaer reclamation area, a cotton field identification model was developed that integrates multi-source satellite remote sensing data with machine learning methods. Using imagery from Sentinel-2, GF-1, and Landsat 8, we performed feature fusion using principal component, Gram–Schmidt (GS), and neural network techniques. Analyses of spectral, vegetation, and texture features revealed that the GS-fused blue bands of Sentinel-2 and Landsat 8 exhibited optimal performance, with a mean value of 16,725, a standard deviation of 2290, and an information entropy of 8.55. These metrics improved by 10,529, 168, and 0.28, respectively, compared with the original Landsat 8 data. In comparative classification experiments, the endmember-based random forest classifier (RFC) achieved the best traditional classification performance, with a kappa value of 0.963 and an overall accuracy (OA) of 97.22% based on 250 samples, resulting in a cotton-field extraction error of 38.58 km². By enhancing the deep learning model, we proposed a U-Net architecture that incorporated a Convolutional Block Attention Module and Atrous Spatial Pyramid Pooling. Using the GS-fused blue band data, the model achieved significantly improved accuracy, with a kappa coefficient of 0.988 and an OA of 98.56%. This advancement reduced the area estimation error to 25.42 km², representing a 34.1% decrease compared with that of the RFC. Based on the optimal model, we constructed a digital map of continuous cotton cropping from 2021 to 2023, which revealed a consistent decline in cotton acreage within the reclaimed areas. This finding underscores the effectiveness of crop rotation policies in mitigating the adverse effects of large-scale monoculture practices. This study confirms that the synergistic integration of multi-source satellite feature fusion and deep learning significantly improves crop identification accuracy, providing reliable technical support for agricultural policy formulation and sustainable farmland management.

Rapid and non-destructive nitrogen diagnosis in fruit orchards is critical for precision fertilization management and crop yield optimization. This study develops and evaluates a practical hyperspectral preprocessing pipeline for leaf nitrogen estimation in Korla fragrant pear (Pyrus sinkiangensis Yü), a commercially important cultivar in southern Xinjiang, China. Hyperspectral reflectance data and corresponding nitrogen measurements were collected from mature leaves of slender-spindle-trained trees. Four preprocessing strategies, comprising multiplicative scatter correction (MSC), wavelet threshold denoising, and their sequential combinations, were systematically compared to assess their effects on spectral information retention and model performance. The successive projections algorithm (SPA) was applied for characteristic wavelength selection, and four regression models, including linear regression (LR), partial least squares regression (PLSR), random forest (RF), and XGBoost, were constructed and evaluated. Results demonstrated that combined preprocessing strategies outperformed single-method approaches, and that preprocessing order significantly influenced predictive accuracy. Nonlinear models consistently outperformed linear models, confirming a pronounced nonlinear relationship between hyperspectral features and leaf nitrogen content. The MSC, followed by wavelet threshold denoising, combined with SPA and XGBoost, achieved the best predictive performance, with R² = 0.754, RMSE = 0.179 mg/g, and RPD = 2.017 on the test set. These findings provide a methodological reference for hyperspectral nitrogen monitoring and preprocessing workflow design under controlled conditions, with potential for further validation in field applications.