| 1. |
- Lindgren, Johan, et al.
(författare)
-
Fossil insect eyes shed light on trilobite optics and the arthropod pigment screen
- 2019
-
Ingår i: Nature. - : Nature Publishing Group. - 0028-0836 .- 1476-4687. ; 573:7772, s. 122-125
-
Tidskriftsartikel (refereegranskat)abstract
- Fossilized eyes permit inferences of the visual capacity of extinct arthropods1–3. However, structural and/or chemical modifications as a result of taphonomic and diagenetic processes can alter the original features, thereby necessitating comparisons with modern species. Here we report the detailed molecular composition and microanatomy of the eyes of 54-million-year-old crane-flies, which together provide a proxy for the interpretation of optical systems in some other ancient arthropods. These well-preserved visual organs comprise calcified corneal lenses that are separated by intervening spaces containing eumelanin pigment. We also show that eumelanin is present in the facet walls of living crane-flies, in which it forms the outermost ommatidial pigment shield in compound eyes incorporating a chitinous cornea. To our knowledge, this is the first record of melanic screening pigments in arthropods, and reveals a fossilization mode in insect eyes that involves a decay-resistant biochrome coupled with early diagenetic mineralization of the ommatidial lenses. The demonstrable secondary calcification of lens cuticle that was initially chitinous has implications for the proposed calcitic corneas of trilobites, which we posit are artefacts of preservation rather than a product of in vivo biomineralization4–7. Although trilobite eyes might have been partly mineralized for mechanical strength, a (more likely) organic composition would have enhanced function via gradient-index optics and increased control of lens shape.
|
|
| 2. |
- Munkevik, Per, et al.
(författare)
-
A computer vision system for appearance-based descriptive sensory evaluation of meals
- 2007
-
Ingår i: Journal of Food Engineering. - : Elsevier BV. - 0260-8774 .- 1873-5770. ; 78:1, s. 246-256
-
Tidskriftsartikel (refereegranskat)abstract
- This paper presents a complete machine vision system for automatic descriptive sensory evaluation of meals. A human sensory panel first developed a set of 72 sensory attributes describing the appearance of a prototypical meal, and then evaluated the intensities of those attributes on a data set of 58 images of example meals. This data was then used both to train and validate the performance of the artificial system. This system covers all stages of image analysis from pre-processing to pattern recognition, including novel techniques for enhancing the segmentation of meal components and extracting image features that mimic the attributes developed by the panel. Artificial neural networks were used to learn the mapping from image features to attribute intensity values. The results showed that the new system was extremely good in learning and reproducing the opinion of the human sensory experts, achieving almost the same performance as the panel members themselves. (c) 2005 Elsevier Ltd. All rights reserved.
|
|
| 3. |
|
|
| 4. |
- Townsend, Philip, 1991, et al.
(författare)
-
Stochastic modelling of 3D fiber structures imaged with X-ray microtomography
- 2021
-
Ingår i: Computational materials science. - : Elsevier B.V.. - 0927-0256 .- 1879-0801. ; 194
-
Tidskriftsartikel (refereegranskat)abstract
- Many products incorporate into their design fibrous material with particular levels of permeability as a way to control the retention and flow of liquid. The production and experimental testing of these materials can be expensive and time consuming, particularly if it needs to be optimised to a desired level of absorbency. We consider a parametric virtual fiber model as a replacement for the real material to facilitate studying the relationship between structure and properties in a cheaper and more convenient manner. 3D image data sets of a sample fibrous material are obtained using X-ray microtomography and the individual fibers isolated. The segmented fibers are used to estimate the parameters of a 3D stochastic model for generating softcore virtual fiber structures. We use several spatial measures to show the consistency between the real and virtual structures, and demonstrate with lattice Boltzmann simulations that our virtual structure has good agreement with respect to the permeability of the physical material. © 2021 The Author(s)
|
|
