Abstract
In recent years, the interest for non-destructive imaging of the internal structures in food products has increased. First of all, the food industry shows an increased interest for automated quality inspection of food products. Secondly, food microstructure has become more important within food science for understanding and designing food products. In both of these aspects, X-ray imaging methods such as radiography and computed tomography provide a non-destructive solution.
However, since the conventional attenuation-based modality suers from poor contrast in soft matter materials, modalities with improved contrast are needed. Two possible candidates in this regard are the novel X-ray phase-contrast and X-ray dark-eld imaging modalities. The contrast in phase-contrast imaging is based on dierences in electron density which is especially useful for soft matter materials whereas dark-eld imaging produces a contrast based on dierences in microstructure.
In order to increase the use of X-ray imaging within food science, possible applications of X-ray phase-contrast and X-ray dark-eld imaging should be studied. To reach these applications, improvements are needed on several aspects of the imaging process. From the initial step of taking the image, the information in the image needs to be translated through image analysis before data analysis can be applied to treat the image quantitatively and answer the questions at hand.
In this work, a number of studies were carried out to investigate possible applications of novel X-ray imaging modalities within food science. The first two studies mainly concern the image acquisition process of taking the image. Using dark-eld radiography, raw, frozen and defrosted fruit were distinguished, and structural changes in barley seeds during germination were monitored. Furthermore, the process of translating the image in image analysis was addressed. For improved handling of multimodal image data, a multivariate segmentation scheme of multimodal X-ray tomography data was implemented. Finally, quantitative data analysis was applied for treating the images. Quantitative studies were conducted on the microstructure of a dairy-like food emulsion as well as the structural changes in meat due to heat treatment.
However, since the conventional attenuation-based modality suers from poor contrast in soft matter materials, modalities with improved contrast are needed. Two possible candidates in this regard are the novel X-ray phase-contrast and X-ray dark-eld imaging modalities. The contrast in phase-contrast imaging is based on dierences in electron density which is especially useful for soft matter materials whereas dark-eld imaging produces a contrast based on dierences in microstructure.
In order to increase the use of X-ray imaging within food science, possible applications of X-ray phase-contrast and X-ray dark-eld imaging should be studied. To reach these applications, improvements are needed on several aspects of the imaging process. From the initial step of taking the image, the information in the image needs to be translated through image analysis before data analysis can be applied to treat the image quantitatively and answer the questions at hand.
In this work, a number of studies were carried out to investigate possible applications of novel X-ray imaging modalities within food science. The first two studies mainly concern the image acquisition process of taking the image. Using dark-eld radiography, raw, frozen and defrosted fruit were distinguished, and structural changes in barley seeds during germination were monitored. Furthermore, the process of translating the image in image analysis was addressed. For improved handling of multimodal image data, a multivariate segmentation scheme of multimodal X-ray tomography data was implemented. Finally, quantitative data analysis was applied for treating the images. Quantitative studies were conducted on the microstructure of a dairy-like food emulsion as well as the structural changes in meat due to heat treatment.
| Originalsprog | Engelsk |
|---|
| Forlag | The Niels Bohr Institute, Faculty of Science, University of Copenhagen |
|---|---|
| Status | Udgivet - 2016 |