Özet:
An approach based on both discrete and continuous wavelet analysis followed by a zero-crossing technique was developed. We applied this approach to obtain a high resolution in the binary mixture of caffeine (CA) and propyphenazone (PR) in the presence of their overlapping signals in the working length. The optimization of the wavelet families was accomplished for this mixture. The de-noise procedure was carried out by using 4-level Haar discrete wavelet transform and the resulted de-noised signal was investigated by continuous Mexican (MEX) and Haar (HA) transforms. Finally, a zero-crossing technique was applied on the transformed signal and the constructed calibration was tested by analyzing the composition of the different mixture containing CA and PR. All calculations have been performed within EXCEL and Matlab 6.5 software. The obtained results indicate that our procedure is flexible and applicable for the mixture analysis.