RESUMO
The fusion of infrared and visible images is a well-researched task in computer vision. These fusion methods create fused images replacing the manual observation of single sensor image, often deployed on edge devices for real-time processing. However, there is an issue of information imbalance between infrared and visible images. Existing methods often fail to emphasize temperature and edge texture information, potentially leading to misinterpretations. Moreover, these methods are computationally complex, and challenging for edge device adaptation. This paper proposes a method that calculates the distribution proportion of infrared pixel values, allocating fusion weights to adaptively highlight key information. It introduces a weight allocation mechanism and MobileBlock with a multispectral information complementary module, innovations which strengthened the model's fusion capabilities, made it more lightweight, and ensured information compensation. Training involves a temperature-color-perception loss function, enabling adaptive weight allocation based on image pair information. Experimental results show superiority over mainstream fusion methods, particularly in the electric power equipment scene and publicly available datasets.
RESUMO
In this paper, a novel 4H-SiC deep-trench super-junction MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) with a split-gate is proposed and theoretically verified by Sentaurus TCAD simulations. A deep trench filled with P-poly-Si combined with the P-SiC region leads to a charge balance effect. Instead of a full-SiC P region in conventional super-junction MOSFET, this new structure reduces the P region in a super-junction MOSFET, thus helping to lower the specific on-resistance. As a result, the figure of merit (FoM, BV2/Ron,sp) of the proposed new structure is 642% and 39.65% higher than the C-MOS and the SJ-MOS, respectively.
RESUMO
Ni35 coatings were fabricated on 45 steel using a CO2 laser at various parameters. A relatively large spot (10 mm diameter) was adopted, which was beneficial to the coating quality and the cladding efficiency. The cross-sectional geometry, phase constituent, and microstructure of the coatings were investigated. With a lower specific energy, coating height increased, while coating width, melted depth, dilution rate, width to height ratio and contact angle decreased. Ni35 coating primarily consisted of γ-Ni, FeNi3, Ni3B, Cr23C6, and Cr5B3. Dendrites with flower-like, fishbone-like, pearl-like, and column-like morphologies were observed. The fraction of flower-like dendrites increased gradually with the decrease in scanning velocity due to the dendrite growth direction evolution. With the decrease in scanning velocity, the microstructure of the heat-affected zone transformed from martensite to martensite + sorbite and finally sorbite. The maximum microhardness of the Ni35 coating reached 451.8 HV0.2, which was about double that of the substrate (220 HV0.2).