F. Antonucci

Printing on Food or Food Printing: a Review

The processes of printing foodstuffs or imaging on an unconventional surface deserve extra attention in comparison with conventional techniques especially in relation with the substrate interested that will or should say a lot about the concept behind the design. Indeed, while printing on food, designers are required to consider carefully both, the appropriate edible inks and the production processes. Moreover, printing on food packaging requires almost as much care. However, the concept of food printing represents nowadays a new frontier in food processing and industry to realize new food products of complex shapes and colour and with particular mixtures. At the moment, many current research projects and products related to food printing are being developed. In 2011, a European Cooperation in Science and Technology action named “New possibilities for print media and packaging, combining print with digital” was created with the aim to promote an interdisciplinary interaction among European research partner with several different research backgrounds. Among the aims of the project, a crucial aspect regards the combination of food expertise with printing using new technologies in order to print on food or food printing. In light of this scenario, image processing and machine control occupy a very important part of the research: a number of programs were written or modified as part of the research into food printing. This review aims to produce an updated analysis on the current developments regarding the technology for food printing and printing on food. In this sense, the work starts giving an overview of the 2D and three-dimensional (3D) printer technology and moves on the food media, edible substrate used by 3D printers and a print of food chapter, related to the substrates used by the 2D printers.

Comparison between manual and stereovision body traits measurements of Lipizzan horses

The accuracy between manual & stereovision measures on Lipizzan horses was measured.The results showed a high total correlation and a low variability between operators.The average stereovision error was < 3% and differences depend by specific traits.The stereovision system prototyped could be a helpful tool for phenotyping. The Lipizzan horse is one of the oldest European horse breeds, its documented origins date back to 1580 in the Imperial stud-farm at Lipizza and are currently used the high-riding school, dressage and others equestrian sports as well as recreational activities. They are also traditionally selected by their morphological characteristics. The aim of this study was to verify the accuracy between manual measurements and stereo-image ones, for the most important linear, angular and circumference measures used for breeding purposes. In this approach a dual web-camera system, in combination with an image analysis algorithm, is proposed to automatically extract the information needed. For this reason a model species such as Lipizzan horse has been chosen and the most important biometric variables (linear measurements and angles) have been selected and taken into consideration. Ten horses were analyzed manually and using a stereovision system taking into account six linear measurements and two angles. The comparison between manual and stereovision measurements showed a high total correlation (r=0.998) and a low variability between operators (SD=0.0004). The average error, lower than 3% and difference in magnitude of error depending by specific traits. In conclusion the stereovision system prototyped could be a good tool to improve phenotyping and enlarge the basis of population involved in breeding programs of horses as well as in other livestock species.

Multivariate approaches to electronic nose and PTR-TOF-MS technologies in agro-food products

Beyond the appearance of agricultural products, affecting the consumers’ choice, volatile organic compounds play a crucial role in agro-industrial processes and all food-related sciences and technologies as well as food perception and acceptability, being at the origin of its aroma and flavor. Thus, this chapter will show the multiple uses of the electronic nose and PTR–TOF–MS technologies in the agricultural field, particularly in postharvest quality control and cultivar classification, mainly focusing on applications regarding citrus fruits and dairy products and showing the potential of multivariate statistical approaches. Moreover, the PTR–TOF–MS, being capable of providing fingerprint for agro-industrial product, opening new promising fields of application in food research, thus providing key elements to improve food production processes and enhance food consumer acceptability.

An Open Source Conveyor Belt Prototype for Image Analysis-Based Rice Yield Determination

The basic role of industrial rice milling is the transformation of paddy rice into white rice with good appearance while selecting the best quality grain for human consumption. In Italy, the commercial value of paddy rice is assessed calculating whole and free of defect kernel yield after processing. The determination is performed by laboratories utilizing a benchtop yield machine that carries out the husking and kernel bleaching. The aim of the study is the development of a pilot conveyor belt (grain coulter), based on image analysis, to increase the reliability of laboratory yield estimation (discrimination of paddy and white grains). The tests regard rice grains belonging to 26 different genotypes of rice grown in Sardinia (Italy). The low-cost prototype based on open source technologies that aim to substitute the current subjective estimation made by eye with an industrial like optically based one. The method is based on the image analysis and extracts three main qualitative attributes: shape, size (i.e., Fourier descriptors and basic morphometry) and appearance (color), and the use of a multivariate classification technique (i.e., partial least squares discriminant analysis). The models discriminated samples of paddy or white rice for each genotype considered (26 models) and for all the genotypes considered together (1 model). For all the 27 models, the mean sensitivities and specificities were very high, ranging from 99 to 100%, while the mean classification errors were very low. The mean percentage of correct classification in the test set was equal to 99.99% for the “unique model” (i.e., paddy VS white rice) and 100% for the 26 single genotype models. The proposed system appears to be useful not only for paddy and white rice discrimination but also as a flexible apparatus for analyzing many other agro-food products. Indeed, the algorithm was used on other food products, such as red hot chili peppers for other discrimination purposes.