Application Of The Optimum Wavelet Packet Transform In Electrophoresis Signals De-noising

Capillary Electrophoresis chip has been recognized as one of the most promising technologies in the field for biochemical analyses. Due to its high sensitivity Laser-induced fluorescence detection method is widely used. Some background noises are introduced into the signal because of the influence of instrument and some other factors, to achieve high sensitivity and accuracy, the de-noising of capillary electrophoresis signal is necessary before data processing.Most of the traditional signal de-noising methods are based on the Fourier transform. But because of the time-frequency unitary of Fourier transform, these methods can not get a good effect when used for the non-stationary signal de-noising. Wavelet transform has the better ability to analyze the singularities and irregular signal because of a multi-resolution analysis, so it is very suitable for being applied to noise on removal capillary electrophoresis electropherograms. Wavelet packet analysis is an extension from the wavelet analysis, which carrying on a more detailed signal analysis and reconstruction.Aiming at the drawbacks of the existing de-noising methods for electrophoretic fluorescent signal, a novel de-noising algorithm based on the best basis of wavelet packets is proposed, it selects the best basis of wavelet packets according to Shannon entropy criteria and uses Penalized threshold. The quantified coefficient of the wavelet decomposition is reconstructed to obtain the de-nosing signal. The results of simulation experiments indicate that this algorithm not only increases the signal-to-noise ratio, but also reduces the peak-error of fluorescent signal.During Wavelet packet de-noising process, the selection of wavelet bases for noise reduction results has a great impact. This is because that different Wavelet bases have different time-frequency characteristics. According to the characteristics of the signal, how to choose the best wavelet base is one of the most critical problems.Aiming at the insufficient for Wavelet bases, the paper presents a design method of matching Wavelet energy. The signal matching wavelet is constructed by the largest projection principle of signal in the scale space, combined with genetic optimization algorithm. In this paper, it describes the process of matching wavelet of the fluorescent signal, and also gives the matching wavelet result. The simulation results show that this Wavelet de-noising method in the wavelet of the performance is better than the ordinary.