An image analysis of TLC patterns for quality control of saffron based on soil salinity effect: A strategy for data (pre)-processing. (15th January 2018)

• Record Type: Journal Article
• Authors: Sereshti, Hassan; Poursorkh, Zahra; Aliakbarzadeh, Ghazaleh; Zarre, Shahin; Ataolahi, Sahar
• Is Part Of: Food chemistry. Volume 239(2018)
• Page Start: 831
• Page End: 839
• Publication Date: 2018-01-15
• Subjects: Image analysis -- Image pre-processing -- Saffron -- Soil salinity -- Thin layer chromatography; Food -- Analysis -- Periodicals; Food -- Composition -- Periodicals; 664
• DOI: 10.1016/j.foodchem.2017.07.012
• Languages: English
• ISSNs: 0308-8146

Abstract

Highlights: Thin layer chromatography-image analysis was used for saffron pattern recognition. The chemometrics data pre-processing was used to remove the TLC images artifacts. PCA and k-means clustering were used to investigate saffron patterns. The effect of soil salinity on the TLC patterns was studied. Abstract: Quality of saffron, a valuable food additive, could considerably affect the consumers' health. In this work, a novel preprocessing strategy for image analysis of saffron thin layer chromatographic (TLC) patterns was introduced. This includes performing a series of image pre-processing techniques on TLC images such as compression, inversion, elimination of general baseline (using asymmetric least squares (AsLS)), removing spots shift and concavity (by correlation optimization warping (COW)), and finally conversion to RGB chromatograms. Subsequently, an unsupervised multivariate data analysis including principal component analysis (PCA) and k-means clustering was utilized to investigate the soil salinity effect, as a cultivation parameter, on saffron TLC patterns. This method was used as a rapid and simple technique to obtain the chemical fingerprints of saffron TLC images. Finally, the separated TLC spots were chemically identified using high-performance liquid chromatography-diode array detection (HPLC-DAD). Accordingly, the saffron quality from different areas of Iran was evaluated and classified.