The application in processing analytical chemistry signals of a cardinal spline approach to wavelets.

The mth order B-spline N(m)(x) with integer knots generates a multiresolution analysis, ···⊂V(-1)⊂V(0)⊂···, with the mth order of approximation in wavelet transformation (WT). We have compared the WT method with Fourier transformation processing analytical chemistry signals in detail and have found that the WT method has many advantages. This method can directly provide the frequency domain distribution of a signal in a time domain. The algorithms are simpler and take less time in operation, and one needs only to decompose the digital signal rather than to undertake many transformations. Moreover, it is not necessary to preprocess the original signal to analyze the pattern of data and to know the statistical character of the noise. When the signal-to-noise ratio is 0.2, the processed results of theoretical and experimental data can still be satisfactory.