刘炜,常庆瑞,郭曼,邢东兴,员永生.小波变换在土壤有机质含量可见/近红外光谱分析中的应用[J].干旱地区农业研究,2010,28(5):241~246
小波变换在土壤有机质含量可见/近红外光谱分析中的应用
Application of wavelet transformation in detection of organic matter contentbased on visible/near infrared reflectance spectroscopy
  
DOI:10.7606/j.issn.1000-7601.2010.05.45
中文关键词:  可见/近红外光谱  土壤有机质  一阶导数光谱  小波变换  定量分析
英文关键词:visible/near infrared spectra  organic matter content  first derivative spectra of soil  wavelet transformation  quantitative analysis
基金项目:“十一五”国家科技支撑计划重大项目(2006BAD09B0603);国家自然科学基金项目(30872073);国家“973”计划项目(2007CB407203)
作者单位
刘炜 西北农林科技大学资源环境学院 陕西 杨凌 712100 
常庆瑞 西北农林科技大学资源环境学院 陕西 杨凌 712100 
郭曼 西北农林科技大学资源环境学院 陕西 杨凌 712100 
邢东兴 西北农林科技大学资源环境学院 陕西 杨凌 712100
咸阳师范学院资源环境系 陕西 咸阳 712000 
员永生 西北农林科技大学资源环境学院 陕西 杨凌 712100 
摘要点击次数: 315
全文下载次数: 152
中文摘要:
      使用高光谱仪ASD Field Spec在波长范围400~1 000 nm内采集有机质含量不同的土壤原始光谱;之后,求取土壤原始光谱及其A值的一阶导数(一阶导数光谱);对以上两类一阶导数光谱进行小波去噪处理;并分析这两类一阶光谱在去噪前、后,光谱形态的变化情况,以及它们与有机质含量相关性的改善程度。结果表明:(1) 土壤原始光谱及其A值的一阶导数中含有大量噪声,严重影响了对光谱曲线波形轮廓、特征信号点和有机质吸收特征的识别。(2) 经小波去噪处理后,土壤原始光谱及其A值的一阶导数中的噪声被有效去除,一阶导数光谱的清晰度明显改善,光谱质量显著提高。(3) 去噪后的土壤原始光谱的一阶导数,呈开口向下的“喇叭口”状,波长范围567~598 nm内与有机质含量呈较为稳定、显著的负相关性。(4) 去噪后的土壤原始光谱A值的一阶导数,波形呈“凹”状,波长范围524~535 nm内与有机质含量呈比较稳定、显著的正相关性。
英文摘要:
      The hyper-spectral reflectance of soil was measured by a ASD FieldSpec within 400~1 000 nm, and processed by logarithmic transformation for getting A value of soil spectra. Next, first order derivative operation was conducted for the original soil spectra and A value. Then, the two types of first derivative spectra of soil above were denoised by the threshold denoising method based on wavelet transformation. In the two kinds of denoised spectra, spectrum response feature due to different soil organic matter content was discussed; and the sensitive bands for organic matter content estimation were initially determined by applying correlation. Results showed that: (1) Because of much noise, it is difficulty to identify contour and response feature in the two types of first derivative spectra resulting from different organic contents. (2) By the threshold denoising method based on wavelet transformation, noise in the two types of first derivative spectra is removed effectively, and the spectra contour and response feature can be identified easily. (3) Within 567~598 nm, there are high and stable negative correlation coefficients between soil organic matter contents and denoised first derivative spectra of soil. (4) Within 524~535 nm, there are high and stable positive correlation coefficients between soil organic matter contents and denoised first derivative A value of soil.
查看全文  查看/发表评论  下载PDF阅读器