Classification Of Pepper Based On Fourier Transform Infrared Spectroscopy

In this dissertation, Fourier transform infrared (FTIR) spectroscopy combinedwith statistical analysis was used to study the Capsicum frutescens L.The results showed that the infrared spectra of Capsicum frutescens L wasmainly composed of the bands of polysaccharides, protein and lipids. The infraredspectra of Capsicum frutescens L were similar, but tiny differences in wave-numbersand absorption intensities of peaks were observed in the range of1800-800cm-1.Fourier transform infrared spectroscopy combined with principal component analysis(PCA) and hierarchical cluster analysis (HCA) were utilized to study the Capsicumfrutescens L cluster. HCA and PCA were performed on the second derivative infraredspectra in the range1800-700cm-1, Hierarchical cluster analysis provides about100%of correct classification. The first three principal components have a cumulativecontribution rate of94.47%, and yield classification accuracies of98%.Fourier transform infrared (FTIR) spectroscopy combined with back propagationneural network (BPNN), wavelet transform and support vector machine (SVM) wereused to study Capsicum frutescens var and bell pepper. FTIR spectra of60sampleswere obtained, the infrared spectra in the range of1750-950cm-1were extracted bycontinuous wavelet transform (CWT) and discrete wavelet transform (DWT). Thedecomposition level20is obvious differences, and three regions of this level areselected as feature vector to train BPNN and the SVM models. Discrete wavelettransform detail coefficients (DWTDC) of level5were selected to train BPNN andSVM models, the recognition accurate rate of BPNN and SVM is93.3%and100%respctively. It is proved that FTIR spectroscopy combined with BPNN and SVMcould be used to discriminate Capsicum frutescens var and bell pepper.Fourier transform infrared (FTIR) spectroscopy combined with principalcomponent analysis (PCA) and partial least squares discriminant analysis (PLS-DA)were used to study Capsicum frutescens var and Capsicum frutescensL. CV. Shuan leanse L. D zhou. H. Liuet. P. H. Licv. Nov. FTIR spectra of60samples of2speciesof Capsicum frutescens L were obtained, the results showed that the infrared spectraof Capsicum frutescens L were similar, but tiny differences in wave-numbers andabsorption intensities of peaks were observed in the range of1800-950cm-1. PCA andPLS-DA were performed on the second derivative infrared spectra in this range.Results show that PCA and PLS-DA can be distinguish two kinds of chili peppers.PCA provides about100%of classification accuracy, and PLS-DA yield86.7%classification accuracy for Capsicum frutescens var and93.3%for Capsicumfrutescens L. CV. Shuan laense L. D Zhou. H. Liuet. P. H. LiCV. NOV. It is provedthat FTIR spectroscopy combined with PCA and PLS-DA could be used todiscriminate Capsicum frutescens var and Capsicum frutescensL. CV. Shuan leanse L.D zhou. H. Liuet. P. H. Licv. Nov.It is proved that FTIR spectroscopy combined with statistical analysis could beused to study the Capsicum frutescens L clusters.