Browsing by Author "Germain, Juan C."

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Identifying industrial food foam structures by 2D surface image analysis and pattern recognition
    (ELSEVIER SCI LTD, 2012) Germain, Juan C.; Aguilera, Jose M.
    Bubbles are fundamental structural elements in several food products modulating density, rheology, texture, appearance and mouthfeel. Foams and aerated structures are characterized by their gas content, stability, bubble size and distribution. However, these measures alone cannot fully describe the complexity of bubble-containing structures. We have used three image analysis methods (Euler characteristic, Minkowski fractal and image texture) to characterize foam structure, and canonical and Bayesian discriminant analysis to identify/classify different foam architectures. This work describes results of this methodology on liquid foams stabilized by proteins at varying concentration and pH levels. Results indicated that groups of three structural parameters (among the 57 calculated) could successfully identify foam structures with different characteristics but unfortunately no single set of features could be used ubiquitously. Additional foam structure information as determined in this work can help to better understand these systems and the impact of bubbles on the physical properties of aerated foods. (C) 2012 Elsevier Ltd. All rights reserved.
  • Thumbnail Image
    Item
    Image Analysis of Representative Food Structures: Application of the Bootstrap Method
    (WILEY, 2009) Ramirez, Cristian; Germain, Juan C.; Aguilera, Jose M.
    Images (for example, photomicrographs) are routinely used as qualitative evidence of the microstructure of foods. In quantitative image analysis it is important to estimate the area (or volume) to be sampled, the field of view, and the resolution. The bootstrap method is proposed to estimate the size of the sampling area as a function of the coefficient of variation (CVBn) and standard error (SEBn) of the bootstrap taking sub-areas of different sizes. The bootstrap method was applied to simulated and real structures (apple tissue). For simulated structures, 10 computer-generated images were constructed containing 225 black circles (elements) and different coefficient of variation (CVimage). For apple tissue, 8 images of apple tissue containing cellular cavities with different CVimage were analyzed. Results confirmed that for simulated and real structures, increasing the size of the sampling area decreased the CVBn and SEBn. Furthermore, there was a linear relationship between the CVimage and CVBn. For example, to obtain a CVBn = 0.10 in an image with CVimage = 0.60, a sampling area of 400 x 400 pixels (11% of whole image) was required, whereas if CVimage = 1.46, a sampling area of 1000 x 100 pixels (69% of whole image) became necessary. This suggests that a large-size dispersion of element sizes in an image requires increasingly larger sampling areas or a larger number of images.