Super‐resolution of the undersampled and subpixel shifted image sequence by a neural network

Numerous approaches to super‐resolution (SR) of sequentially observed images (image sequence) of low resolution (LR) have been presented in the past two decades. However, neural network methods are almost ignored for solving SR problems. This is because the SR problem traditionally has been regarded as the optimization of an ill‐posed large set of linear equations. A designed neural network based on this has a large number of neurons, thereby requiring a long learning time. Also, the deduced cost function is overly complex. These defects limit applications of a neural network to an SR problem. We think that the underlying meaning of the SR problem should refer to super‐resolving an imaging system by image sequence observation, instead of merely improving the image sequence itself. SR can be regarded as a pattern mapping from LR to SR images. The parameters of the pattern mapping can be learned from the imaging process of the image sequence. This article presents a neural network for SR based on learning from the imaging process of the image sequence. In order to speed up the convergence, we employ vector mapping to train the neural network. A mapping vector is composed of some neighbor subpixels. Such a well‐trained neural network has powerful generalization ability so that it can be used directly to estimate the SR image of the other image sequences without learning again. Our simulations show the effectiveness of the proposed neural network. © 2004 Wiley Periodicals, Inc. Int J Imaging Syst Technol 14, 8–15, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.20001     

Assuring Microstructural Homegeniety in Dual Phase and Trip Steels

The presence of ferrite/pearlite bands in dual phase and TRIP assisted steels is a consequence of microchemical segregation which causes mechanical properties anisotropy. Such inhomogeneous phase distribution produces a lowering of the mechanical properties such as fracture behaviour. This anisotropy is commonly not accounted in micromechanics computations which often assume a random distribution of phases in the solid. The present paper deals with an integral model for this undesirable band formation accounting for the solute segregation caused by solidification, microcomponent diffusion present in the austenitisation process, and the nucleation of the transformed phase in segregated regions. In the present work, the model was applied to two industrial grade dual phase steels and two TRIP assisted steels. The influence of such parameters on band formation is summarised in a number of “band prevention plots”, which are aimed at providing the optimum processing conditions for ferrite/pearlite band prevention.     

Predicting cetane number,kinematic viscosity,density and higher heating value of biodiesel from its fatty acid methyl ester composition

Biodiesel is a renewable bio-fuel derived from natural fats or vegetable oils, and it is considered as a promising alternative to substitute diesel fuels. Cetane number, viscosity, density, and higher heating value are important properties to affect the utilization of biodiesel fuels, because they are involved in the definition of fuel quality and are required as input data for predictive engine combustion models. This work presents the characterization of two biodiesel samples made from beef tallow and soybean oil through their fatty acid methyl esters (FAMEs) profile. Empirical equations were developed to estimate four physical properties of methyl esters; and an average absolute deviation (AAD) of 5.95%, 2.57%, 0.11% and 0.21% for the cetane number, kinematic viscosity, density, and higher heating value were founded. Cetane number, viscosity, and higher heating value increases because of the increase of molecular weight and these physical properties decrease as the number of double bonds increases. Unlike that of above properties, density decreases as molecular weight increases and density increases as the degree of unsaturation increases. Two general mixing rules and five biodiesel samples were used to study the influence of FAMEs over the physical properties of biodiesel. The prediction of the cetane number, kinematic viscosity, density and higher heating value of biodiesel is very close to the experimental values.     

Lipolysis,lipid peroxidation and texture of Serrano ham processed under different ripening temperature conditions

Lipolysis, lipid peroxidation, texture and rancid taste were investigated in Serrano hams processed under low, medium and high ripening temperature conditions (RTC) for 15 months. Medium RTC hams showed the highest contents of saturated and monounsaturated free fatty acids (FFAs) from month 5 to 15 and of polyunsaturated FFAs from month 7 to 12. The primary peroxidation index decreased during ripening in all hams, with higher levels for low RTC hams from month 5 onwards. Contrarily, the secondary peroxidation index increased during ripening in all hams, with higher levels for medium RTC hams from month 7 onwards. Texture parameters varied significantly among ham muscles. Shear force increased during ripening in all hams, with higher values for medium RTC hams, whereas cutting force was not influenced by RTC or ripening time. Rancid taste scarcely developed during ripening and was not affected by RTC.     

Dispersive Liquid–Liquid Microextraction Followed by Capillary High-Performance Liquid Chromatography for the Determination of Six Sulfonylurea Herbicides in Fruit Juices

In this study a simple, rapid, and efficient method has been developed for the determination of six sulfonylurea herbicides (SUHs): triasulfuron, metsulfuron-methyl, chlorsulfuron, flazasulfuron, chlorimuron-ethyl, and primisulfuron-methyl in commercial grape and apple juice samples, using dispersive liquid–liquid microextraction coupled with capillary high-performance liquid chromatography with diode array detection. Various parameters that influence the extraction efficiency, such as the type and volume of extraction and disperser solvents, sample pH, and salt addition, were investigated and optimized. Under the optimum conditions, limits of detection and quantification of the method were in the ranges of 2–9 and 8–29 μg L^?1, respectively, lower than the maximum residue limits set by the European Union for the raw fruits, such as grape and apple. The intra- and inter-day relative standard deviations varied from 1.0 to 8.2 and 1.8 to 9.8 %, respectively, with recoveries between 72.0 and 109.5 % for commercial grape (both white and red) and apple juice samples, showing satisfactory accuracy for the determination of SUHs in fruit juices.     

Technological Implications of Modifying the Extent of Cell Wall-Proanthocyanidin Interactions Using Enzymes

The transference and reactivity of proanthocyanidins is an important issue that affects the technological processing of some fruits, such as grapes and apples. These processes are affected by proanthocyanidins bound to cell wall polysaccharides, which are present in high concentrations during the processing of the fruits. Therefore, the effective extraction of proanthocyanidins from fruits to their juices or derived products will depend on the ability to manage these associations, and, in this respect, enzymes that degrade these polysaccharides could play an important role. The main objective of this work was to test the role of pure hydrolytic enzymes (polygalacturonase and cellulose) and a commercial enzyme containing these two activities on the extent of proanthocyanidin-cell wall interactions. The results showed that the modification promoted by enzymes reduced the amount of proanthocyanidins adsorbed to cell walls since they contributed to the degradation and release of the cell wall polysaccharides, which diffused into the model solution. Some of these released polysaccharides also presented some reactivity towards the proanthocyanidins present in a model solution.     

Design of aperiodic planar arrays for isoflux radiation in GEO satellites by applying evolutionary optimization

The design of an aperiodic planar array is presented in this research. This design of aperiodic arrays considers the reduction of the side lobe level and the isoflux radiation requirements for GEO (Geostationary Earth Orbit) satellite applications. In this way, it is considered four different optimization cases. The first two cases are optimizations of amplitude and phase excitations for the antenna elements in a uniform antenna array and the last two cases are optimizations of positions of the antenna elements and certain number of levels of amplitude excitation in an aperiodic array. In this case, it is proposed a simple new approach combining the main idea of both thinned theory and random arrays approaches. For this optimization problem, the well-known method of Genetic Algorithms (GA) is utilized. The obtained results show the proper performance of the array factor to provide the isoflux radiation and low side lobe level. Depending on the performance requirements, the design of the aperiodic array could lead the satellite hardware to be reduced significantly even more that results presented previously in the literature.     

In vitro digestibility of phenolic compounds from edible fruits: could it be explained by chemometrics?

The health benefits of phenolic compounds depend on the ingested amount, molecular diversity and gastrointestinal digestibility. The phenolic profile of eight fruits (blackberry, blueberry, strawberry, raspberry, mulberry, pomegranate, green and red globe grapes) was chemometrically associated with their in vitro digestibility (oral, gastric, intestinal). Extractable phenols, flavonoids and anthocyanins strongly correlated with each other (r ≥ 0.84), proanthocyanidins with anthocyanins (r = 0.62) and hydrolysable phenols with both extractable phenols (r = 0.45) and proanthocyanidins (r = ?0.54). Two principal components explained 93% of the variance [61% (free‐phenols), 32% (bounded‐phenols)], and four clusters were confirmed by hierarchical analysis, based in their phenolic richness (CLT 1‐4: low to high) and molecular diversity. In vitro digestibility of extractable phenols and flavonoids was blackberry (CLT‐4)> raspberry (CLT‐2)> red grape (CLT‐1) related to their phenolic richness (r ≥ 0.96; P < 0.001), but anthocyanins’ digestibility was pH‐dependent. Chemometrics is useful to predict the in vitro digestibility of phenolic compounds in the assayed fruits.