02923 Statistical and physical modeling for visual food inspection
Statistisk og fysisk modellering til visuel fødevareinspektion
Point( ECTS )
Taught under open university
We will schedule 15 gatherings in the period from September 2012 to May 2013
Scope and form:
Seminar series where students present scientific papers.
Duration of Course:
[The course is not following DTUs normal Schedule]
Type of assessment:
General course objectives:
The purpose of this course is that the participants learn techniques for visual food inspection at the PhD level. The appearance of food items can be related to food quality parameters, and advanced statistical tools are needed for inferring these properties. Another important element is to understand the physical light scattering that determines the appearance of the food items. During the course we will read literature covering the topics of food inspection, advanced statistical tools, and physical modeling of light scattering. We plan to have 10 two hours meetings, where each student will get a presentation assignment of about 20 minutes such that the meetings cover the literature. The remaining time at the meetings will be spent discussing the literature of the day.
Overall, this course serves to act as an inspiration for research within visual food inspection and improve the level of competency of the participating PhD students.
A student who has met the objectives of the course will be able to:
Use advanced statistical inference models including sparse features selection, classification and regression models.
Use advanced image acquisition techniques for food inspection.
Use theory of light scattering, including Lorenz-Mie theory, to connect food properties and contents to the appearance of food items.
Select relevant physical models for simulating light transport in biological samples.
Select important image features for food inspection and determine how to obtain these features from images.
Use hyperspectral techniques for food inspection.
Identify important food quality parameters that can be obtained from images.
Use state-of-the-art techniques for computer vision based food inspection.
The course is build around scientific papers. We will use the literature listed below, or similar papers within the scope of the course.
Hastie, T., Tibshirani, R., Friedman, J., and Franklin, J. (2009): The elements of statistical learning: data mining, inference and prediction. Springer. Selected chapters.
Sun, D. W. (2010): Hyperspectral imaging for food quality analysis and control. Academic Press/Elsevier. Selected chapters.
Jensen, H. W., Marschner, S. R., Levoy, M., and Hanrahan, P. (2001): A practical model for subsurface light transport. Proceedings of ACM SIGGRAPH 2001, pp. 511-518.
Gowen, A. A., O'Donnell, C. P., Cullen, P. J., Downey, G., and Frias, J. M. (2007): Hyperspectral imaging - an emerging process analytical tool for food quality and safety control. Trends in Food Science & Technology 18(12), pp. 590-598.
Frisvad, J. R., Christensen, N. J., and Jensen H. W. (2009): Predicting the appearance of materials using Lorenz-Mie theory. Unpublished manuscript.
Hielscher, A. H., Mourant, J. R., and Bigio, I. J. (1997): Influence of particle size on the diffuse backscattering of polarized light from tissue phantoms and biological cell suspensions. Applied Optics 36(1), pp. 125-135.
Kim, Y. L., Liu, Y., Wali, R. K., Roy, H. K., Goldberg, M. J., Kromin, A. K., Chen, K., and Backman, V. (2003): Simultaneous measurement of angular and spectral properties of light scattering for characterization of tissue microarchitecture and its alterations in early precancer. IEEE Journal of Selected Topics in Quantum Electronics 9(2), pp. 243-256.
, Building 324, room 113, Ph. (+45) 4525 3415
, Building 324, room 220, Ph. (+45) 4525 3764
, Building 321, room 210, Ph. (+45) 4525 3358
01 Department of Applied Mathematics and Computer Science
Registration Sign up:
Last updated: 02. maj, 2013