摘要
红外光谱在采集过程中,由于外界环境条件以及仪器的影响,采集谱图会出现一定程度的基线漂移。为消除漂移基线的影响,提出了一种改进的非对称最小二乘基线校正算法。算法在非对称最小二乘基础上加入一阶导数约束项,并利用直方图背景估计方法的计算值作为基线的初始值,通过迭代地求解,得到最终的基线。所得基线平滑性好,并能保证在原始谱图下方。仿真和真实光谱数据实验表明,所提出的算法在保持峰位和峰形不变的同时,有效地消除了谱图的基线,取得了较好的效果。
Due to the effects of spectrometer and the change of environmental conditions, the measured IR spectra often contains undesirable baseline artifacts. An improved asymmetric least squares baseline correction algorithm is proposed. A first-order derivative constraint term is incorporated in the algorithm. It uses the background values as the initial baseline, and solves the optimization problem iteratively. The obtained baseline is very smooth and always under the original spectrum. Experimental results on both simulated data and real data demonstrate the proposed algorithm can perform effectively in eliminating the baseline drift while preserving the position and shape of the peaks in the original spectra.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2012年第5期537-540,共4页
Computers and Applied Chemistry
基金
国家自然科学基金资助项目(61101219
71141020
61032007)
关键词
红外光谱
基线漂移
非对称最小二乘
基线校正
Infrared spectrum, Baseline drift, Asymmetric least squares, Baseline correction
