基于连续小波特征的植烟土壤有机碳含量反演研究

打开文本图片集

中图分类号：S572；S126 文献标识码：A文章编号：1007-5119（2025）05-0108-07

Inversion of Soil Organic Carbon Content in Tobacco Planting Soil Based on Continuous Wavelet Analysis

ZHAO Ben1， WANG Peng²，LI Beibei1， WANG Tao³， WU Yong²， ZHANG Qian1， WUMing²，LI Zhigang²，YE Xiefengl

（1.Tobacco ScienceCollegeofHenanAgriculturalUniversity/NationalTobacoCutivationPhysiologicalandBiochemicaleseachBase/KeyLaboratoryofTbaccCultivationinobaccoIndustryZhengzou45046，China;2.WuhanCigareteFactoryofCina Tobacco Hubei Tobaco Industrial Limited Liability Company， Wuhan 430oo， China; 3.Qujing Branch ofYunnan Tobacco Company， Qujing 655ooo， Yunnan， China）

Abstract：Toachieveaccrate inversionofsoilorganiccarbon （SOC）content intobacco-growingsoilunderlimitedsensitivebands， this studyutized106soil samples from three counties in Qujing Cityas research subjects.Aninversion model was usedto predict SOCcontentin tobaco-growingsoil basedonthecombinationof sensitive bands and their wavelet features.A model forinverting SOCcontent based oncontinuous wavelet analysis was established usingahyperspectral imaging device.Results indicated that the regression modelestablished withthe hyperspectral data optimized bySavitzky-Golayflteringshowed anaverage increaseof 44% in thecoeficietofeteiaionomparedtoatbeforetetiatin.Usingteelatiooientetod，itwasfoudtat 733nm was the optimal spectral band for SOC in tobacco planting，with a coefficient of determination （R2） value of 0.75. By applying continuouswavelettransfortoteoptiizedhyperspectralreflectancedata，itwasobseedthattheiversionaccracyinceased with wavelet features at scales from1 to 10 under the 733nm band， stabilizing after scale 6， where the R2 value exceeded 0.8.Four machine learning models were developed using the optimized 733nm spectral data combined with its wavelet features extracted at scales 6 to 10 as input variables. Among these， the Random Forest model produced the best prediction results， with R2 values of 0.88， root mean square errors of 1.65g/kg ，relative prediction deviations of3，and running time of O.21 s，respectively.Combining sensitive bands with theirownhigh-scale waveletfeatures eablesrapid，non-destructive，andaccurate iversioofSOCcontent intobac growing soil. These results provide a reference for the development of algorithms for SOC sensors.

Keywords： tobacco planting soil; hyperspectral reflectance; wavelet analysis; soil organic carbo

土壤有机碳（soilorganiccarbon，SOC）是土壤有机质的主要组成部分，是评价土壤质量的重要指标之一。（剩余12697字）