Imagens digitais e térmicas, espectroscopia NIR e calibração multivariada na avaliação de parâmetros de textura de farinhas

Abstract

Wheat flour is a predominant ingredient in several food products and there has been a gradual insertion into the consumer market of various flours and with specific parameters of nutritional and technological quality for food production, such as whole wheat flour, corn, corn white, cassava starch, oats, rye, rice, brown rice, brown linseed,golden linseed, soybeans, bananas, sweet potatoes, passion fruit peel, passion fruit, eggplant, ginger, yacon, amaranth, chia, coconut, grapes that can be substituted partial or whole wheat, thus ensuring a healthier life for the population that needs gluten-free flour, and each type of flour has specific particularities related to texture, fractureability, hardness, adhesiveness, elasticity, resilience, cohesiveness, gumminess and chewability. The objective of this work was to use parameters of digital images, thermal images, spectroscopy in the region of near infrared (NIR) and to use the chemometric tool, multivariate calibration (PLS) in the investigation of correlation with parameters of texture analyzed in the texturometer (TPA). In this case, different types of flour were used in the same multivariate calibration model, it is considered multiproduct. The models obtained from the digital images presented higher adjustment results, the models generated from the thermal images and from the NIR spectroscopy also present satisfactory results. The results of VIP scores allow us to verify that hardness, gumminess and cohesion are similar in the variables important to model these parameters. It was concluded that the proposed method is an alternative to the conventional texturometer method for determining texture parameters in flours.