摘要
东北典型黑土区是保证粮食安全与生态稳定的核心区域,近年来面临日益严重的土壤退化问题,急需通过遥感技术提供空间信息更丰富的数据。文中针对黑土区光谱特征复杂、遥感数据分辨率不足的问题,开展地块、区域、景观多尺度影像融合方法对其质量进行评价,以提升地物分类与土壤属性反演精度。本研究以松嫩平原中部黑土区作为实验区域,采用CN、Brovey、Gram-Schmidt、HSV、PC、NNDiffuse共6种方法对Landsat8全色和多光谱数据进行融合,并采用多个指标对融合效果进行客观评价。结果表明,在地块与区域尺度下NNDiffuse方法在信息量与空间细节上表现最优,景观尺度下PC方法效果最佳。光谱信息保持度方面,不同尺度下NNDiffuse方法均显著优于其他方法。
The typical black soil region in Northeast China is a core area for safeguarding food security and ecological stability.In recent years,it has been confronted with increasingly severe soil degradation,urgently requiring remote sensing technology to provide data with richer spatial information.Addressing the issues of complex spectral characteristics and insufficient resolution of remote sensing data in the black soil region,this study explores multi-scale image fusion methods(plot,regional,and landscape scales)and evaluates their quality,aiming to improve the accuracy of land cover classification and soil attribute inversion.Taking the black soil area in the central Songnen Plain as the experimental region,six methods(CN,Brovey,Gram-Schmidt,HSV,PC,and NNDiffuse)were applied to fuse Landsat 8 panchromatic and multispectral data,with multiple indicators used for objective evaluation of fusion effects.Results show that at the plot and regional scales,the NNDiffuse method performs best in information content and spatial details,while the PC method achieves optimal results at the landscape scale.In terms of spectral information retention,the NNDiffuse method is significantly superior to other methods at all scales.
作者
赵景琦
李丹
席志龙
李长春
ZHAO Jingqi;LI Dan;XI Zhilong;LI Changchun(College of Surveying and Mapping Engineering,Heilongjiang Institute of Technology,Harbin 150050,China;School of Surveying and Land Information Engineering,Henan Polytechnic University,Henan Jiaozuo 454002,China)
出处
《测绘工程》
2025年第6期56-63,共8页
Engineering of Surveying and Mapping
基金
2023年度黑龙江省省属高等学校基本科研业务费科研项目(2023GJ03)
黑龙江省自然科学基金(PL2024D018)
2024年度黑龙江省省属本科高校基本科研业务费科研项目(2024BJ03)。
