Determination of tomato quality with hyperspectral imaging
¹Latvia University of Life Sciences and Technologies, Faculty of Agriculture,
Institute of Soil and Plant Sciences, Liela Str. 2, LV-3001 Jelgava, Latvia
²Latvia University of Life Sciences and Technologies, Faculty of Food Technology, Department of Chemistry, Liela Str. 2, LV-3001 Jelgava, Latvia
³Latvia University of Life Sciences and Technologies, Faculty of Information Technologies, Department of Mathematics, Liela Str 2, LV-3001 Jelgava, Latvia
⁴Institute of Vegetable and Melon Growing, Department of Breeding and Seed Production of Vegetable and Melon Crops, Instytutska Str. 1, Selektsiyne village, UA68478 Kharkiv region, Ukraine
⁵Riga Technical University, Faculty of Power and Electrical Engineering, Kalku Str. 1, LV-1048 Riga, Latvia
*Correspondence: ina.alsina@lbtu.lv
Abstract:
Tomatoes (Solanum lycopersicum L.) are a widely used vegetable in the human diet throughout the year, both fresh and in various processed products. Tomatoes contain compounds important to human health and are an important source of vitamins, antioxidants, and mineral elements. Performing biochemical analyses is an expensive, environmentally unfriendly and time-consuming process; therefore, a way to determine the biochemical composition of tomatoes using non-destructive methods is being sought. The study includes 45 varieties of tomatoes with different colors – red, pink, orange, brown, yellow, and bicolor tomato fruits. The content of dry matter, soluble dry matter, titratable acidity, lycopene, β-carotene, total phenol, and flavonoids was determined by standard biochemical procedure. Reflectance spectrums of tomato fruits were obtained with Remote Sensing Portable Spectroradiometer RS-3500 (Ltd. Spectral Evolution, Haverhill, MA, USA) at the wavelength 350–2,500 nm with a 1 nm interval. In order to determine the content of various biochemical parameters in tomatoes, the vegetation indices found in the literature were used, and new ones were developed. The research demonstrated that the developed vegetative indices allow to detect lycopene and β-carotene content non-destructively. For the determination of the dry matter, soluble solids and phenolic content, indices designed for detecting water content can be used, but their correlation coefficients with chemical methods are moderately high – 0.65, 0.56 and 0.57, respectively. It was found that the best correlation between biochemically detected parameters and vegetation indices is for lycopene > β-carotene > dry matter> total phenols = titratable acidity ≥ soluble solids > taste index > flavonoids.
Key words:
lycopene, taste index, total phenols, Vis-NIR, β-carotene