Reading handwritten digits: a ZIP code recognition system

A neural network algorithm-based system that reads handwritten ZIP codes appearing on real US mail is described. The system uses a recognition-based segmenter, that is a hybrid of connected-components analysis (CCA), vertical cuts, and a neural network recognizer. Connected components that are single digits are handled by CCA. CCs that are combined or dissected digits are handled by the vertical-cut segmenter. The four main stages of processing are preprocessing, in which noise is removed and the digits are deslanted, CCA segmentation and recognition, vertical-cut-point estimation and segmentation, and directly lookup. The system was trained and tested on approximately 10000 images, five- and nine-digit ZIP code fields taken from real mail     

Investigation of physical and mechanical properties of oil palm wood core sandwich panels overlaid with a rubberwood veneer face

Low-density sandwich panels consisting of an oil palm wood core overlaid with a rubberwood veneer face were manufactured. Effects of two types of grain orientation of the oil palm wood core (parallel and perpendicular to board surface) and three different veneer thicknesses (0.7, 1.8 and 2.7 mm) and core densities (223 ± 14, 301 ± 35 and 418 ± 33 kg/m³) on some physical and mechanical properties of the boards were investigated. Results showed that higher core density increased the values of thermal conductivity, screw withdrawal resistance, modulus of rupture and modulus of elasticity but decreased the value of water absorption without effect on thickness swelling of the boards. Boards with the core grain direction oriented perpendicular to panel’s surface possessed lower value of thickness swelling but higher values of thermal conductivity and strain at fracture when the board failed in a mode of core shear under bending test than those of the others. Finally, the relationship between board density and the measured physical and mechanical properties of the oil palm wood core sandwich panels overlaid with a rubberwood veneer expressed as mathematical equations could be used to predict and design the expected properties of this type of sandwich board.     

Decision-making and opinion formation in simple networks

In many networked decision-making settings, information about the world is distributed across multiple agents and agents’ success depends on their ability to aggregate and reason about their local information over time. This paper presents a computational model of information aggregation in such settings in which agents’ utilities depend on an unknown event. Agents initially receive a noisy signal about the event and take actions repeatedly while observing the actions of their neighbors in the network at each round. Such settings characterize many distributed systems such as sensor networks for intrusion detection and routing systems for Internet traffic. Using the model, we show that (1) agents converge in action and in knowledge for a general class of decision-making rules and for all network structures; (2) all networks converge to playing the same action regardless of the network structure; and (3) for particular network configurations, agents can converge to the correct action when using a well-defined class of myopic decision rules. These theoretical results are also supported by a new simulation-based open-source empirical test-bed for facilitating the study of information aggregation in general networks.     

Reaction‐Bonded Boron Carbide/Magnesium–Silicon Composites

Reaction‐bonded boron carbide was manufactured by infiltrating porous boron carbide preforms at 1273 K with a Mg‐Si eutectic alloy. The resulting composite material consists, in addition to the original B₄C, of SiC, Mg₂Si, and a Mg‐rich complex boride/carbide Mg_x(Al,Si)_y(B,C)_z phase. The composites display high hardness (1700 HV), Young's modulus (356 MPa) and a moderate bending strength (230 MPa). The ballistic efficiency (of about 6.7), as determined by the depth of penetration method, is much higher than that of alumina and similar to that of silicon‐infiltrated reaction‐bonded composites.     

Within-tree variability of internal stress generated during drying of rubberwood lumber

The generation of drying stress within the lumber from the trunk of a rubber tree prepared from different locations (radial distance up to 110 mm and height up to 4 m) has been investigated in real-time by using a restoring force measurement on half-split specimens. Drying was performed at constant dry-bulb and wet-bulb temperatures of 90 and 60?°C, respectively. The entire restoring force profiles do not vary significantly with height. In addition, before and after the reversal of stress, the force profiles are largely similar and the maximum negative and positive forces are approximately equal regardless of wood locations within the tree trunk. However in the radial direction, the process of stress reversal consisting of two negative force maxima appears to proceed slower in the inner juvenile wood than in the outer mature wood. Upon water immersion of the specimens for 4 months, the second negative force maximum gradually disappears and the force profiles with a shorter stress reversal period become less sensitive to the wood locations. An examination of the drying curves in the second stage of drying during stress reversal also indicates a slower migration of bound water out of the lumber in the juvenile wood compared to that in the mature wood. The drying is also faster in the water-immersed specimens. It is concluded that variability of the internal stress within the trunk of a rubber tree originated from the role of cell wall amorphous constituents and cell wall extractives on creep property and the movement of bound water within the wood cell wall during drying.     

Antifungal Activity of Crude Extracts of Coffee and Spent Coffee Ground on Areca Palm Leaf Sheath (Areca catechu) Based Food Packaging

This study observed the antifungal activity of crude coffee extract (CCE) and crude spent coffee ground extract (CSCGE) against Aspergillus sp., Aspergillus flavus, Aspergillus niger, Penicillum citrinum, Penicillum chrysogenum and Eurotium amstelodami on the surface of an areca palm leaf sheath (Areca catechu). The agar dilution method was employed to determine the minimal inhibitory concentration (MIC) of the CCE for the first, second and third extractions and CSCGE for the first and second extractions in malt extract agar at concentrations from 50 to 500 μg/ml. A mould test was then performed on durian paste product wrapped with the areca palm leaf sheath treated with the second CSCGE at its MIC. Gas chromatography‐mass spectrometry (GC‐MS) was performed to find the major components of the CCE and CSCGE and to measure the total phenolic content. For the second CSCGE treatment on the areca palm leaf sheath, the following qualities were studied: mould growth, wettability, measurements of hardness and thickness and a Fourier transform infrared spectroscopy scan. It was found that the CCE and CSCGE MICs from the first and second extractions ranged from 100 to 230 μg/ml and from 300 to 460 μg/ml, respectively. In addition, the MICs showed the capability of CCE and CSCGE to provide protection against mould growth on the leaf sheath for at least 21 days in comparison with the control (3 days) during storage at accelerated conditions (25 °C and 100% relative humidity). Shelf life of durian paste product wrapped with areca palm leaf sheath treated with the second CSCGE at 460 μg/ml was extended from 3 days to 21 days under storage condition of 30 °C. Caffeine was the main constituent of CCE (79.69%) and CSCGE (84.92%). Also, the total phenolic content of CCE and CSCGE measured were 0.640 and 0.981 mol ferulic acid 100 g^?1, respectively. This study has demonstrated that the first and second CCE and CSCGE were capable of inhibiting mould growth under both in vitro and in vivo tests. Copyright © 2015 John Wiley & Sons, Ltd.     

The Diverse Transformer (Trf) Protein Family in the Sea Urchin Paracentrotus lividus Acts through a Collaboration between Cellular and Humoral Immune Effector Arms

Sea urchins are long-living marine invertebrates with a complex innate immune system, which includes expanded families of immune receptors. A central immune gene family in sea urchins encodes the Transformer (Trf) proteins. The Trf family has been studied mainly in the purple sea urchin Strongylocentrotus purpuratus. Here, we explore this protein family in the Mediterranean Sea urchin Paracentrotus lividus. The PlTrf genes and predicted proteins are highly diverse and show a typical Trf size range and structure. Coelomocytes and cell-free coelomic fluid from P. lividus contain different PlTrf protein repertoires with a shared subset, that bind specifically to E. coli. Using FACS, we identified five different P. lividus coelomocyte sub-populations with cell surface PlTrf protein expression. The relative abundance of the PlTrf-positive cells increases sharply following immune challenge with E. coli, but not following challenge with LPS or the sea urchin pathogen, Vibrio penaeicida. Phagocytosis of E. coli by P. lividus phagocytes is mediated through the cell-free coelomic fluid and is inhibited by blocking PlTrf activity with anti-SpTrf antibodies. Together, our results suggest a collaboration between cellular and humoral PlTrf-mediated effector arms in the P. lividus specific immune response to pathogens.