Moisture-content is one of the most significant properties to determine quality of carrot during storage. The optical measurement methods of moisture content promise non-destructive, non-contact and fast solution for quality control, for monitoring quality changes during storage and also for real-time classification tasks.The high absorption coefficient of water makes NIR analysers a commonly used tool for accurate moisture analysis. Hyperspectral system is able to detect the spatial distribution of reflectance spectrum as well. In case of finding correlation between the moisture-content of carrot and the reflectance spectral data, a hyperspectral system would be suitable for testing quality.Experiments were made to investigate spectral changes of different cultivars and different tissues of carrot stored under different conditions. Moisture-decrease of pieces and also the spectral data of carrot slices were recorded. Statistical analysis of the data has shown the optimal intensity function to describe moisture-content. Eliminating homogeneous spectral changes caused by destructed tissues, only a narrow interval of NIR range was sensitive to the moisture-decrease of xylem tissues.The equipment and the measurement procedure were able to identify carrot tissues and detect their changes during drying. For non-destructive applications of the system, further experiments are needed to inspect the behaviour of rhizodermis.
Authors:V. Parrag, J. Felföldi, L. Baranyai, A. Geösel, and F. Firtha
From the nineteen-nineties, cobweb disease caused serious losses for the mushroom sector in Europe, in the USA, and in Australia (Fletcher & Gaze, 2008), so it is one of the most notable fungal infections of cultivated white button mushroom (Agaricus bisporus). The aim of this study was to identify cobweb disease (Cladobortyum dendroides) caused cap spotting and brownish rot on the mushroom sporocarp, and to find a proper discrimination method in the case of this infection.Fruiting body samples were divided into 4 groups, a control one and three others treated with different chemicals that are tested against fungal infections. The groups were subdivided into 2 portions and the first was infected with cobweb disease. Images of the caps were recorded and their hyperspectral images were acquired in the wavelength range of 900–1700 nm.On the hyperspectral images infected and healthy areas were selected, on these average spectra differences were found around the known water peaks (1200 and 1450 nm). The spatial distribution of the water content can be used for the detection of the spoilage, because the infected areas showed different reflection values at these water absorption peaks.Support Vector Machine method was applied successfully to discriminate between the infected and control groups and Monte Carlo cross-validation was carried out.
Authors:J. Felfoldi, L. Baranyai, F. Firtha, L. Friedrich, and Cs. Balla
The fat content (fat distribution) of the pork and beef raw material is one of their most important quality characteristics. Image processing methods were applied to provide with quantitative parameters related to these properties. Different hardware tools were tested to select the appropriate imaging alternative. Statistical analysis of the RGB data was performed in order to find appropriate classification function for segmentation. Discriminant analysis of the RGB data of selected image regions (fat-meat-background) resulted in a good segmentation of the fat regions. Classification function was applied on the RGB images of the samples, to identify and measure the regions in question. The fat-meat ratio and textural parameters (entropy, contrast, etc.) were determined. Comparison of the image parameters with the sensory evaluation results showed an encouraging correlation.
This study was carried out to evaluate some physical and mechanical properties of three Hungarian rice varieties named Dáma, Janka and M488 under different moisture levels to be a useful data for modelling the moisture changes in rough rice storage bins. Rough rice grains were conditioned to moisture contents of 12, 18, 24, and 30% (w.b.). Five mechanical expressions named rupture force (Fr), maximum stress (σmax), grain deformation (D), energy (E) and toughness (T) were extracted from stress–strain curve for all tested varieties as a function of moisture content. Also, some physical properties such as bulk density, true density, porosity and some morphological features of grain were determined as a function of moisture content for tested rough rice grain. The results revealed that the measured physical and mechanical properties are significantly effected by moisture contents for the three tested rice varieties. In general, when the moisture content increased, the rupture force and maximum stress decreased for all investigated grains. However, the deformation, energy and toughness firstly decreased with moisture content increase and after that increased with further increase of moisture content for all rice varieties. On the other hand, the bulk density, true density and porosity do not have a specific trend with increasing moisture content. Moreover, there was a significant difference among the selected rice varieties in terms of their bulk density and porosity at the same moisture content range.
Authors:K. Szalay, J. Deákvári, F. Firtha, I. Tolner, Á. Csorba, and L. Fenyvesi
The hyperspectral imaging spectroscopy is a promising future tool in the field of optical remote sensing and it creates new perspective for modern information management in site specific agricultural production. One can determine quantitative relationships between the environmental and physiological parameters of vegetation cover and the soil quality parameters as well as the features of the reflectance spectra by the newgeneration data monitoring and sampling method. These reflectance spectra have characteristics of the different crops and provide with the possibility of accurate classification and detection. The objective was to present the technological capabilities of hyperspectral imaging and show some exprimental results of nutrient sensitive changes in the winter wheat spectra. There were found two characteristic wavelength ranges: the 500 to 800 nm for wheat kernel samples and the 1650 nm to 1800 nm for wheat ear samples where fertilizer treatments showed definite trend on the basis of the normalized reflectance spectra.
Authors:V. Parrag, Z. Gillay, Z. Kovács, A. Zitek, K. Böhm, B. Hinterstoisser, R. Krska, M. Sulyok, J. Felföldi, F. Firtha, and L. Baranyai
One of the most important food safety issues is the detection of mycotoxins, the ubiquitous, natural contaminants in cereals. Hyperspectral imaging (HSI) is a new method in food science, it can be used to predict non-destructively the changes in composition and distribution of compounds. That is why, in the last decade, the potential of HSI has been evaluated in many fields of food science, including mycotoxin research.
The aim of the recent study was to test the feasibility of HSI for the differentiation according to the toxin content of cornmeal samples inoculated with Fusarium graminearum, Fusarium verticillioides and Fusarium culmorum and samples with natural levels of mycotoxins. Samples were measured in the near infrared wavelength range of 900–1,700 nm and mean spectra of selected regions of interest of each image were pre-treated using Savitzky-Golay smoothing and standard normal variate (SNV) method. On the spectra, partial least squares discriminant analysis (PLS-DA) was carried out according to the level of contamination. Partial least squares regression (PLSR) method was used to predict deoxynivalenol (DON) content of samples and the cumulative toxin content: the sum of fumonisins (FB1, FB2) and DON content of samples. Based on the promising results of the study, HSI has the potential to be used as a preliminary testing method for mycotoxin content in feed materials.