Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In the agro-food industry, there is an increasing demand for non-destructive, fast and cost-efficient methods for the objective determination of product quality. Several parameters influence the quality of a product, while the importance of each of these parameters may vary amongst people. Optical measurement techniques are often used because of their non-destructive nature. However, the robust determination of both chemical and physical quality attributes remains difficult. In this work, a non-destructive determination of quality was performed using laser scatter imaging. This technique allows to obtain more information on both the absorption and scattering of light, by retrieving spatial information. The bulk optical properties (bulk absorption coefficient, bulk scattering coefficient and scattering anisotropy factor) are used to characterize absorption and scattering properties.^ The absorption of light is related to the chemical composition of a product, like present pigments, water or sugars. Scattering of light is more related to the physical structure, possibly allowing to retrieve more information on physical quality attributes, like firmness, tenderness or porosity. Optical properties of light were derived from the obtained scatter images using a data-based modelling technique, possibly allowing a better prediction of product quality. This approach was tested in two case studies, apples and bovine meat, selected because of their economic importance for Belgium.First, a hyperspectral laser scatter imaging (HLSI) system was developed. A combination of a supercontinuum laser with monochromator was used to illuminate samples with monochromatic light in the 550 nm to 1000 nm range, while a CCD camera was used to take images of the diffuse reflectance glow spots.^ From this glow spots, a diffuse reflectance profile with the light intensity relative to the distance from the point illumination was obtained. Models were constructed to optimize the detector size and the source-detector distance, estimating different quality parameters of Braeburn apples. A detector size of 0.82 mm was found to be adequate for the estimation of all parameters, including the starch value, firmness, SSC and Streif index. Different source-detector distances were found to be of importance for predicting different quality traits. Photons exiting the sample closer to the point illumination, which have interacted less with the sample, were more important for SSC prediction, while the prediction of physical parameters like firmness relied on photons which had more interaction with the sample. Moreover, using variable selection, the wavelength regions of pigment and water absorption were found to be most informative.^ A classification of apples into ripeness classes based on these models was possible, with a misclassification rate of 12.5%. Nevertheless, these models still used mixed information, including both the effects of absorption and scattering.Using the double integrating spheres (DIS), the golden standard method for determining bulk optical properties (BOP), the interaction of light with both apple and bovine meat samples was studied. Both the apple skin and cortex were studied separately during maturation of bi-colored (Braeburn and Kanzi) and green cultivars (Greenstar). The bulk absorption coefficient µa of the skin showed features of anthocyanins at 550 nm, chlorophyll at 678 nm and water at 970 nm, 1200 nm and 1450 nm, while the µa of the cortex showed an overall lower absorption attributed to carotenoids, chlorophyll and water.^ During maturation of apples, an increase in the absorption by anthocyanins was observed in the red cultivar’s skin, while a decrease in absorption by chlorophyll was observed in the cortex. Both the bulk scattering coefficient µs and anisotropy factor g of the skin were significantly higher in comparison to the cortex. Both skin and cortex were found to be highly forward scattering with anisotropy factors above 0.9 in the entire wavelength range between 500 nm and 1850 nm. During maturation, no clear evolutions in the anisotropy factor were observed, while µs decreased in the fruit cortex. It was hypothesized that the shape and size of the scattering particles hardly changed during maturation.The DIS analysis showed changes in the optical properties of apple during maturation. This indicates that the non-destructive estimation of BOP could be beneficial in determining apple quality.^ To go from scatter images towards an estimation of the optical properties, a data-based model was used. To train this model, a set of optical phantoms with known optical properties was measured using the HLSI system. These diffuse reflectance measurements, in combination with the BOP from the DIS, were used as an input for a metamodel, linking the BOP to diffuse reflectance profiles. The metamodel showed a good performance for a set of validation phantoms, with an R2V of 0.9977 and 0.957 in combination with an RMSEV of 0.20 cm-1 and 3.21 cm-1, for respectively µa and µs’. Nevertheless, at wavelengths with extreme BOP values, the predictions were less accurate. The prediction of apple BOP showed an expected course for the µa spectra, with absorption features of anthocyanin, chlorophyll and water. However, an incomplete separation between µa and µs’ was obtained, as µs’ still showed some distinct absorption features.^ Nevertheless, at wavelengths with low absorption, the estimation of µs’ was according to expectations. In addition, the same evolutions in BOP as found with the DIS setup were also found with the non-destructive HLSI technique. Though, no clear relation was obtained between apple quality and the estimated BOP. Possibly, the low variability of both SSC and firmness during maturation, together with a high variability amongst apples from the same harvest day, could have complicated the prediction models. Moreover, still mostly information on the changes of apple pigments was used in the models.Next, two bovine muscles were measured using the DIS as well. Both the longissimus lumborum (LL) and the biceps femoris (BF) were considered, while the BF was further divided into an outer and inner part, due to two-toning. Clear absorption features of myoglobin, mainly oxymyoglobin at 544 nm and 582 nm, and water were found in the µa spectra.^ A higher absorption of myoglobin was found in the BF samples, while also a higher µs and a lower g was found in this muscle compared to the LL. The inner and outer BF showed significant scattering differences, possibly related to an increased degree of protein denaturation in the inner BF. During wet aging, when meat tenderness increases, a decrease in the measured g was noticed in both muscle samples.When measuring muscle samples using spatially resolved spectroscopy, anisotropic light scattering by the present muscle fibers should be accounted for. By measuring muscle samples with different initial fiber orientations, it was shown that muscle fibers can change the shape of the diffuse reflectance glow spots from circular to a rhombus shape. This effect was mainly present in the samples with the muscle fibers running parallel to the measurement surface. Moreover, the rhombus’ major axis orientation was related to the distance from the point of illumination.^ Close to the point illumination the major axis orientation was found to be perpendicular to the muscle fibers, while at larger distance a 90° shift occurred, aligning the major axis with the muscle fibers. In these samples with muscle fibers parallel to the measurement surface, the fiber orientation could be predicted based on the fitted rhombuses, with an R2P of 0.993 and RMSEP of 3.95°. These results show the importance of the 3D fiber orientation when measuring diffuse reflectance signals. Moreover, this 3D fiber orientation could possibly be determined using the obtained diffuse reflectance signals.The prediction of muscle BOP values using the metamodel also showed similarities with the DIS measurement. Clear absorption features of oxymyoglobin and water were present, while the absorption by metmyoglobin was observed as well, related to the ticker samples measured using HLSI.^ A thicker sample allows a gradient of myoglobin forms to exist, relative to different depths inside the sample. Again, larger µa and µs’ values were observed for the inner BF samples, while both the outer BF and LL samples showed lower values for µs’. During wet aging, a significant increase was obtained for the µs’ of the LL muscle. Moreover, due to the non-destructive nature of HLSI, measurements were possible through the plastic vacuum pack. Due to the lack of oxygen inside the vacuum pack, mainly deoxymyoglobin with an absorption peak around 557 nm was present. Measurements on the exact same sample during wet aging, through the vacuum pack, also showed an increase in µs’ of the LL muscle.^ As the changes in meat tenderness were most prominent in the LL muscle, it was suggested that the increase in meat tenderness could explain the observed increase in µs’ for this muscle.Finally, a limited number of wavelengths were selected to design a multispectral hand-held measurement device. Four laser diode modules emitting at wavelengths of 533.3 nm, 674 nm, 800.7 nm and 981.1 nm, were selected and mounted around a CCD camera. Using shutters, the laser light of the different modules was guided onto the sample sequentially. Again, a metamodel was built by measuring a set of liquid optical phantoms with a wide range of both µa and µs’. In validation, the metamodel showed an R2V of 0.9724 and 0.9377 in combination with an RMSEV of 0.56 cm-1 and 5.13 cm-1 for µa and µs’ respectively.^ However, for the prediction of apple and pear samples, with and without the skin, an incomplete separation between absorption and scattering properties was obtained, mainly at wavelengths with high absorption values. Nevertheless, the est

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Cows milk and derived products are important components in the Western diet. The composition of raw milk and the size distribution of suspended fat globules and casein micelles largely determine the nutritional, physicochemical and functional characteristics of the milk. Moreover, as these properties have a considerable impact on the general perception of the milk and the properties of derived products, they are important quality parameters for the dairy industry and the consumer. Additionally, the composition of the extracted raw milk and the size distribution of the suspended particles comprise valuable information on the metabolic, nutritional and general health status of the cow as there is a strong interaction between the cows blood circulation and the milk production. Therefore, techniques to measure the milk quality can help to improve cow health monitoring and enhance the efficiency and animal welfare on a dairy farm.^ Further on, the use of similar sensing-technology in a dairy plant could enable online monitoring of production processes to further reduce the variability and promote the development of new food products with improved properties.Optical measurement techniques are frequently used to monitor processes in industry and agriculture. Among them, spectroscopy studies the interaction between electromagnetic radiation and the product of interest. As the molecules of the product absorb this electromagnetic radiation at specific wavelengths, spectral analysis gives insight in the composition of the measured product. The molecular absorption of visible (Vis) and near-infrared (NIR) radiation is sufficient to be used for the analysis of the main sample components, while the penetration depth of Vis/NIR radiation is still adequate to reduce the need for extensive sample preparation. Nowadays, spectral detectors for this wavelength range are fast, robust and cost-efficient.^ Hence, it is widely used for non-destructive and online quality control in industry, food processing and agriculture. Several researchers have shown the potential of Vis/NIR spectroscopic analysis of raw milk in the lab. However, recent studies indicate that this technology might not be robust against changing scattering properties that are typical for on farm raw milk samples. Scattering is a process where the radiation is forced to deviate from a straight trajectory due to local non-uniformities. In milk, the fat globules and casein micelles are the main cause of Vis/NIR light scattering. This increases the travelling path of the Vis/NIR radiation to an unknown extent and seriously complicates the prediction of the sample composition from measured spectra. Accordingly, powerful and advanced techniques are needed to separate and extract pure absorption and scattering information from the obtained spectra.^ The pure absorption spectrum relates directly to the sample composition according to the Beer-Lambert law. Accordingly, prediction of the milk composition from the pure absorption spectrum with multivariate statistics would be more robust and independent of the effects of light scattering. The scattering properties, on the other hand, are determined by the physical microstructure properties (particle size distribution) of the sample. For milk, this primarily relates to the quantity and size of the suspended fat globules and, to a smaller extent, the casein micelles. Therefore, these scattering properties could be employed to extract this microstructure information, creating an added value for Vis/NIR spectroscopy on milk.Knowledge on the Vis/NIR optical properties of milk is essential for the design and optimization of a measurement configuration and model that would result in accurate and robust estimations for the composition and microstructure of milk.^ Accordingly, the measurement and study of these properties is the main objective of this dissertation. To this end, two sub-objectives were set: (1) the development of a measurement setup for accurate optical characterization of turbid media in the Vis/NIR wavelength range; and (2) to study the effect of quantity and size distribution of the suspended particles on the scattering properties of milk.The full optical characterization of turbid media is not straightforward and can only be achieved through an indirect method where multiple measurements, reflectance and/or transmittance, are combined with an inverted theoretical light propagation model. The samples total reflectance and total transmittance, as measured with double integrating spheres, together with an unscattered transmittance measurement are generally accepted as the golden standard method to estimate bulk optical properties (bulk absorption coefficient, bulk scattering coefficient and scattering anisotropy factor).^ Therefore, a dedicated measurement setup was designed and built to acquire these measurements with high signal-to-noise ratios for turbid and absorbing media in the Vis/NIR wavelength range. This setup consists of a flexible high-power light source, which produces a pre-dispersed narrow collimated light beam and allows for fast and automated wavelength and waveband selection, two integrating spheres with detectors and an unscattered transmittance measurement path. The bulk optical properties of the sample can be extracted from these measurements with an inverse adding-doubling algorithm adapted from literature and optimized for the setup. The measurement and estimation procedure to obtain the bulk optical properties for turbid media was thoroughly validated on a set of 57 liquid optical phantoms.^ The phantom set was designed to cover a wide range of absorption and scattering properties by mixing intralipid (scattering agent), methylene-blue (absorption agent) and water (dilution agent) in different ratios, similar to the phantoms often used to validate measurement systems in the field of biomedical optics. Intralipid is an oil-in-water emulsion which is, except for the much smaller fat globule size, very similar to raw milk. It was found that the followed approach resulted in very accurate estimation of the samples pure absorption and scattering propertiesThe obtained dataset was further explored to investigate the effect of an increasing concentration of scattering intralipid particles on the phantoms scattering properties. Furthermore, it was tested when the particle density is so high that individual scattering events start to influence each other, a phenomenon known as dependent or correlated scattering.^ It was found that dependent scattering has a significant impact on the scattering properties of intralipid-dilutions for particle volume concentrations above 2%. Additionally, semi-empirical equations were derived, describing the scattering properties as a function of the radiation wavelength and the volume concentration of scattering intralipid particles, taking into account dependent scattering.To study the effect of the size distribution of suspended spherical particles on the samples scattering properties, a simulation algorithm was developed that relates them. A generalization of the Mie solution for Maxwells equations was used to calculate the optical properties for a single spherical and scattering particle in an absorbing host medium. Accordingly, the optical properties were combined for multiple particles and polydispersity was supported by discretization of the provided particle size distribution.^ The number of discrete intervals is optimized automatically in an efficient iterative procedure. Finally, the developed microscale light propagation algorithm was validated by simulating the bulk optical properties for two aqueous nanoparticle systems and intralipid in the Vis/NIR wavelength range, taking into account the representative particle sizes. The simulated bulk optical properties matched closely with those obtained by the golden standard method.In a next step, the gathered knowledge, measurement techniques and models were employed to study the effect of the fat globule size distribution on the Vis/NIR bulk scattering properties of milk. Ultrasonic homogenization was performed on raw milk to create milk samples with different fat globule size distributions. Next, the Vis/NIR total reflectance and total and unscattered transmittance spectra of these samples were measured and their bulk optical properties were estimated as described earlier.^ Additionally, the actual particle size distribution of fat globules and casein micelles was measured for each sample and the obtained distribution was used as an input for the microscale algorithm, described earlier, to simulate the samples bulk optical properties. Consequently, the validity of the developed microscale algorithm could be tested for milk by comparing the measured and simulated bulk optical properties. The simulated values were very close to the measured ones as long as scattering was independent. Moreover, it was found that a reduction in the fat globule size results in a higher wavelength-dependency of both the Vis/NIR bulk scattering coefficient and the scattering anisotropy factor. Therefore, these scattering parameters are very suitable to estimate the fat globule size distribution from.^ However, this requires inversion of the microscale model.Finally, the Vis/NIR bulk optical properties were determined for a diverse set of 60 raw milk samples that are representative for individual milkings on a dairy farm. The observed variability was discussed and the relation between the obtained bulk optical properties and the raw milk composition and fat globule size was extensively studied. The bulk absorption coefficient spectra were found to mainly contain information on the water, milk fat and casein content, while the bulk scattering coefficient spectra turned out to be primarily influenced by the quantity and the size of the fat globules. Moreover, there was a strong positive correlation between the fat content in raw milk and the measured

Academic collection --- Theses

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The quality of fruit and vegetables is very important to producers, retailers and consumers. The consumer decides, based on the observed quality of the product, whether or not to buy the product. Next to the flavour and the texture of the fruit, also the appearance is a critical quality parameter. Nowadays, inspection of this appearance is achieved through visual inspection, although it is subjective, not very precise and prone to human errors. Therefore, an automation of this quality inspection would be beneficial. A promising technique for this automation is hyperspectral imaging, enabling the measurement of spectral information in all the pixels in an image.Although many researchers have demonstrated the added value of hyperspectral imaging for quality inspection of agrofood products, this technology was still rarely used in the agrofood industry at the start of this PhD research. One of the reasons was the acquisition time of spectral hypercubes which was typically high. However, this acquisition speed is considerably improved by companies like IMEC, that recently developed hyperspectral cameras with a high acquisition speed. The main reason was the large added v needed to build calibration models and the limited added value to justify the price. Therefore, in this PhD research, the focus was on improving the flexibility of hyperspectral imaging and, as the colour and colour distribution of fruit or vegetables is an important parameter to determine its quality and ripeness, on the development of a method to determine the colour of tomatoes in every pixel. In this way, a fast and contact-free method could be obtained. These concepts were elaborated for vine tomatoes, which were used as a model case containing different challenges, as they have a curved and very glossy surface and vine tomatoes consist of different components, which can vary in quality considerably.First, a semi-supervised algorithm was developed to segment hyperspectral images. It consists of three steps. First, several unsupervised algorithms are tested to split a spectral hypercube in a predefined number of comparable classes. By an operator, based on a visual inspection of the segmentation, the best split is determined. Based on this selection, from each class, a number of pixels is selected using a selection algorithm, which are used as input to develop a supervised segmentation model. This supervised model is used to segment newly measured spectral hypercubes of the same cultivar. Spectra of badly segmented spectral hypercubes are then used to augment the training set and to obtain a more robust model. After training with 10 new spectral hypercubes, the best result was achieved using a Partial Least Squares – Discriminant Analysis (PLS-DA), using a first derivative preprocessing. Applied on 5 additional spectral hypercubes, an overall accuracy on pixel level of 96.95 % for vine tomatoes and 98.52 % for table grapes was achieved. After the initial segmentation of spectral hypercubes in its present classes, the quality of fruit and vegetables can be determined.Another important quality parameter of vine tomatoes is the ripeness. Ripeness of tomatoes is linked to the colour, as during ripening the concentration of chlorophyll decreases while the concentration of lycopene increases, which results in a colour change from green to red. To measure the average colour and the variability present in each tomato, the colour of each pixel in the segmented tomatoes was determined. Therefore, two different methods are compared. The first method is by using the calculations developed by CIE to determine the L*a*b*-values. This method was suited for flat, matte samples, but in the case of vine tomatoes, which are curved and very glossy, the results were not accurate. Therefore, a databased method is presented. This method was suited to determine the hue-angle (R² = 0.95), the a*-value (R² = 0.93) and the L*-value (R² = 0.86) accurately. However, the disadvantage of this method is that it can only be used to determine the colour of comparable tomatoes.Next, it was investigated if the colour of a batch of tomatoes during storage could be predicted based on a measurement shortly after harvest. First, a model was built to describe the variability of a batch of tomatoes during storage. Therefore, mixed effects modelling was used. To describe the colour evolution accurately, 2 random effects were needed. As an analytical solution in this case is very difficult, a data-driven approach was developed. By using the algorithm to determine the hue of tomatoes, the hue in every pixel of the tomato in the image could be accurately determined. Next, a distribution was fitted over all the results of each tomato. The mean and skewness of this distribution was determined and used, together with the time after harvest and the time at which the colour should be known, as input in a multiple linear regression. This made it possible to predict the hue of tomatoes at a certain moment in the near future, until 10 days after harvest, with an R² of more than 0.80 and a RMSE of less than 9°.High quality products need to be free of defects. As a large variability in surface defects can occur, training a method to detect all possible defects is very difficult. Therefore, an algorithm was developed to detect defects on tomatoes by training a model to know good quality tomatoes. Defects are detected as outlying from this good quality. Resulting from an analysis of the spectra of the different tomatoes, the wavelength range between 700 nm and 985 nm was selected and an area-normalization was used as preprocessing. Then, the preprocessed spectra of each individual tomato was used as input of a Principal Component Analysis (PCA), so defective areas could be detected based on a combination of Hotelling’s T²-values and Q-residuals, evaluated by comparing a local difference against a threshold value. After optimisation of this threshold, the algorithm is able to detect puncture damage, but the detection of bruises and cutting damage is more difficult. It could be observed that specular reflections had a large influence on the detection result. By using a cross-polarized hyperspectral setup, it was possible to reduce this effect, reducing the numbers of false-positive detections.The algorithms described above are an important step towards a more easy, intuitive application of hyperspectral imaging. To improve the results achieved by the defect detection algorithm, the initial segmentation should be optimised. It is important that the number of misclassifications resulting from the semi-supervised segmentation algorithm is as low as possible, as these are an important source of misclassifications during the defect detection. Next, methods which are able to account for the curvature of the products and the presence of specular reflections should be developed, as they lead to low accuracies achieved when measuring the colour using the equations described by CIE. A possible solution that should be investigated, is the application of the cross-polarised illumination to reduce the effect of specular reflections on the colour calculations. To ensure that the developed algorithms are robust, they should also be tested on different products, like grapes or cranberries.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The quality of emulsions and suspensions is not only depending on their chemical composition, but also the size of the colloidal particles is a crucial factor. The particle size distribution (PSD) has an impact on the stability and the technical properties, like viscosity, of the product. For many products for example in the food industry, not only the basis product is, but also the processing and derivatives are influenced by the PSD. For example in the case of milk, the ease of creaming and the moisture retention in cheese depend on the size of the milk fat globules. The PSD is also an important factor concerning health: the efficiency of nutrient uptake in the gastrointestinal tract depends on the particle size.Given the impact of the particle size distribution on the quality of the product, the PSD should be accurately monitored during before, during and after the production process. A lot of the techniques currently used for particle size measurements require dilution of dense emulsions and suspensions that are characterised by high light scattering. Moreover, the particle size distribution is often measured on a sample that is assumed to be representative for the entire batch of the product. Replacing such off-line methods by accurate optical in-line PSD measurements, would allow a faster detection of deviation in produced particle size and thus a faster adjustment of the production process if needed.The goal of this research is to develop an accurate method for particle size distribution determination of emulsions and suspensions, by combining spectroscopic measurements with light transport models. The sensor design will be optimised for so that even for dense systems an accurate PSD estimation can be given without the need for sample preparation.To start, an inverse micro-scale light transport model will be constructed based on analytic approximations of Mie theory. The model describes the relationship between the light scattering properties of the medium and the particle size distribution. First, this invers model will be elaborated for samples with a low volume concentration of scattering particles, i.e. samples for which the assumption of independent light scattering is valid. Secondly, this model will be extended to a to an inverse model that takes into account dependent scattering. Such model allows PSD estimation for dense systems with a high volume concentration of scattering particles. This inverse model to estimate particle size distributions from optical properties will be used in the insilico optimisation of a sensor design. By means of light transport simulation, the configuration of a spatially resolved reflectance sensor will be optimised for PSD measurements.For the validation of results, a model system with well-known optical properties is used: suspensions of silica particles in water. Once the models and algorithms are validated on this system, they will be applied to two case studies from the agro-food industry: milk and sauces. These two systems can be seen as models for large variety of emulsions and suspensions from different sectors such as the agro-food industry, pharmacy, paint production,...