Three-dimension fluorescence spectrum compression based on wavelet transform-principal components analysis

Three dimensional fluorescence spectrum data of mineral oil is huge and mixes with a lot of irrelevant information, i.e. redundancy and noise, to solve this problem, using principal component analysis (PCA) with wavelet transform (WT) to compress three dimensional fluorescence spectrum data eliminating irrelevant information at the greatest degree. First, using WT to eliminate irrelevant information, and then, complete compression using PCA. Wavelet transform make up for the insufficiency of PCA in eliminating high-order correlation. Through lots of experiments, fixed the best parameters of wavelet transform, then applied in 10 kinds of spectrum data, at the same variance accumulation contribution rate (e.g. 98%), the data compression ratio is improved and the main characteristic information is well retained. The experimental results show that the wavelet transform combined with principal component analysis method is better suitable for mineral oil'3D fluorescence spectrum compression.