分形理论在光谱识别中的应用

分形理论是研究一类不规则、混乱复杂，但其局部和整体具有相似性体系的科学。分形维数是分形理论中用于描述对象的不规则度和自相似性的基本度量。文章以符合朗伯-比尔定律的光谱信号为研究对象，在概述分形几何基本原理的基础上，提出了以分形维数作为光谱识别特征的方法，运用相空间重构得出了光谱信号的分形维数，通过对光谱信号的分形维数进行比较，达到识别不同光谱的目的，最后举例对该方法进行了说明。

Application of Fractal Theory to Spectral Signal Recognition

The fractal theory is a discipline that studies a kind of irregular and chaotic object with similarity of its part and whole.The fractal dimension is a basic index mark of irregularity and self-similarity in the fractal theory.The present paper takes the spectral signal according with Lambert-beer' law as the object,introduces the basic theory of fractal geometry in brief,puts(forward) the method of fractal dimension as the feature of spectral signal recognition,makes use of reconstructing phase space to gain the fractal dimension of spectral signal,compares different values of the fractal dimension to recognize different spectral signal,and gives an example for explanation.