The Studies on the Choice of the Best Wavelet in Compressing IR by Wavelet Transformation

Li Menglong, Kang bin#, Qi Huayi, Wen Zhining
(Department of Chemistry, Sichuan University, Chengdu 610064)
(#College of Light Industry and Food Engineering , Sichuan University, Chengdu 610065)

Abstract This paper studied the influence of different type of wavelets on compressibility of infrared (IR) spectra. First, five representative IR spectra were selected from Aldrich library of FTIR spectra “Aldrich Condensed Phase”. Then four different wavelet families such as Daubechies wavelets, Coiflets wavelets, Symlets wavelets and biorthogonal wavelets at different order are selected for investigation. The results indicated that Daubechies wavelets which are often employed in the literatures do not hold the best compressibility, but biorthogonal wavelets do it, of which bior 4.4 and bior2.8 hold the best general compressibility.
Key words Compressibility, IR spectra, Type of wavelet function, Wavelet transform

　

红外光谱小波压缩中最佳小波函数的选取

李梦龙 康 彬# 戚华溢 文志宁
（四川大学化学学院 成都 610064 #四川大学轻工与食品学院 成都 610065）

摘要   本文研究了不同小波函数系对红外谱图压缩性能的影响。通过对Aldrich凝聚相傅立叶红外谱图库中10607张谱图的考察，把整个谱图库分为五类，分别选取一个最具代表意义的红外光谱，对Daubechies小波函数系、 Coiflets 小波函数系、Symlets小波函数系和双正交小波函数系在不同的小波函数阶数下进行实验。研究结果表明：一般文献中常用的Daubechies小波函数系并不具有最优压缩性能，双正交小波函数系才是最佳小波函数，其中最优综合压缩性能的小波函数是bior4.4和bior2.8。
关键词  压缩性能  红外光谱  小波函数类型   小波变换

李梦龙 男，39岁，博士生导师，从事化学计量学方面研究。 E-mail:liml@scu.edu.cn
国家自然科学基金资助项目(29877016)
2001-09-24收稿，2002-04-01修回

(PDF全文下载c02070)