Soret and Dufour effects between two rectangular plane walls with heat source/sink

This study addresses the thermo‐diffusion and the diffusion‐thermo phenomena in a semi‐infinite absorbent channel whose walls are contracting/expanding, with heat source/sink effects. The governing partial differential equations with suitable boundary conditions are transformed to a system of dimensionless ordinary differential equations. An analytic solution of the problem has been found using a technique called homotopy analysis method (HAM). HAM gives consistently valid answers to the problem over an extensive variety of parameters and also provides better accuracy. To validate the analytical results, a comparison has been presented with a numerical solution calculated by using the parallel shooting method. The effects of dimensionless parameters, that is, deformation parameter, Reynolds number, Soret and Dufour numbers, and heat source/sink parameter on the expressions of velocity, temperature, and concentration profiles are analyzed graphically to understand the physics of the deformable channel. It has been noted that the velocity across the channel is higher for the expanding channel, as compared to that for the contracting channel. Also the Soret and Dufour number increases the temperature of the fluid, and decreases the concentration. The temperature profile has an increasing behavior in the case of heat source, and a decreasing behavior in the case of heat sink.     

Manipulating Deformable Linear Objects: Fuzzy-Based Active Vibration Damping Skill

Human can handle a deformable object and damp its vibration with recognized skill. However, for an industrial robot, handling a deformable object with acute vibration is often a difficult task. This paper addresses the problem of active damping skill for handling deformable linear objects (DLOs) by using a strategy inspired from human manipulation skills. The strategy is illustrated by several rules, which are explained by a fuzzy and a P controller. A proportional-integral-derivative (PID) controller is also employed to explain the rules as a comparison. The interpretations from controllers are translated into high level commands in a robotic language V+. A standard industrial robot with a force/torque sensor mounted on the wrist was employed to demonstrate the skill. Experimental results showed the fuzzy based damping skill is quite effective and stable even without any previous acknowledge of the deformable linear objects.Category (5)     

ASRNet: Adversarial Segmentation and Registration Networks for Multispectral Fundus Images

Multispectral imaging (MSI) technique is often used to capture images of the fundus by illuminating it with different wavelengths of light. However, these images are taken at different points in time such that eyeball movements can cause misalignment between consecutive images. The multispectral image sequence reveals important information in the form of retinal and choroidal blood vessel maps, which can help ophthalmologists to analyze the morphology of these blood vessels in detail. This in turn can lead to a high diagnostic accuracy of several diseases. In this paper, we propose a novel semi-supervised end-to-end deep learning framework called “Adversarial Segmentation and Registration Nets” (ASRNet) for the simultaneous estimation of the blood vessel segmentation and the registration of multispectral images via an adversarial learning process. ASRNet consists of two subnetworks: (i) A segmentation module S that fulfills the blood vessel segmentation task, and (ii) A registration module R that estimates the spatial correspondence of an image pair. Based on the segmention-driven registration network, we train the segmentation network using a semi-supervised adversarial learning strategy. Our experimental results show that the proposed ASRNet can achieve state-of-the-art accuracy in segmentation and registration tasks performed with real MSI datasets.     

蓄水条件下某水电工程坝址区变形体变形 与稳定性研究

库岸边坡在蓄水后的变形和稳定问题一直是水电工程的主要问题, 尤其是坝址区库岸边坡稳定性对水电工程 的正常运营起着至关重要的控制作用。针对黄河上游某水电站库岸一变形体边坡, 在对该变形体边坡基本特征分 析的基础上, 采用数值计算的手段, 对黄河上游某水电站坝址区变形体在自然及蓄水条件下稳定性进行对比分析。 结果表明: 蓄水后变形体边坡的变形具有明显的分区特征, 且坡体的变形受影响较大。同时, 利用强度折减法对变 形体的稳定性系数进行了计算, 变形体的稳定性较好, 考虑到坡体局部稳定性较差, 可能会产生局部失稳破坏, 因 此, 需采取支护措施保证该处坝址区变形体在运营期间的长期稳定性。     

Continuous Self‐Collision Detection for Deformable Surfaces Interacting with Solid Models

In this paper, we propose a new continuous self‐collision detection (CSCD) method for a deformable surface that interacts with a simple solid model. The method is developed based on the radial‐view‐based culling method. Our method is suitable for the deformable surface that has large contact region with the solid model. The deformable surface may consist of small round‐shaped holes. At the pre‐processing stage, the holes of the deformable surface are filled with ghost triangles so as to make the mesh of the deformable surface watertight. An observer primitive (i.e. a point or a line segment) is computed so that it lies inside the solid model. At the runtime stage, the orientations of triangles with respect to the observer primitive are evaluated. The collision status of the deformable surface is then determined. We evaluated our method for several animations including virtual garments. Experimental results show that our method improves the process of CSCD.     

Robust Treatment of Degenerate Elements in Interactive Corotational FEM Simulations

We address the problem of robust and efficient treatment of element collapse and inversion in corotational FEM simulations of deformable objects in two and three dimensions, and show that existing degeneration treatment methods have previously unreported flaws that seriously threaten robustness and physical plausibility in interactive applications. We propose a new method that avoids such flaws, yields faster and smoother degeneration recovery and extends the range of well‐behaved degenerate configurations without adding significant complexity or computational cost to standard explicit and quasi‐implicit solvers.     

改进R-FCN模型的小尺度行人检测

目的 为了有效解决传统行人检测算法在分辨率低、行人尺寸较小等情境下检测精度低的问题,将基于区域全卷积网络（region-based fully convolutional networks,R-FCN）的目标检测算法引入到行人检测中,提出一种改进R-FCN模型的小尺度行人检测算法。方法 为了使特征提取更加准确,在ResNet-101的conv5阶段中嵌入可变形卷积层,扩大特征图的感受野;为提高小尺寸行人检测精度,在ResNet-101中增加另一条检测路径,对不同尺寸大小的特征图进行感兴趣区域池化;为解决小尺寸行人检测中的误检问题,利用自举策略的非极大值抑制算法代替传统的非极大值抑制算法。结果 在基准数据集Caltech上进行评估,实验表明,改进的R-FCN算法与具有代表性的单阶段检测器（single shot multiBox detector,SSD）算法和两阶段检测器中的Faster R-CNN（region convolutional neural network）算法相比,检测精度分别提高了3.29%和2.78%;在相同ResNet-101基础网络下,检测精度比原始R-FCN算法提高了12.10%。结论 本文提出的改进R-FCN模型,使小尺寸行人检测精度更加准确。相比原始模型,改进的R-FCN模型对行人检测的精确率和召回率有更好的平衡能力,在保证精确率的同时,具有更大的召回率。     

基于改进HRNet的眼底视网膜血管分割算法

针对现有眼底视网膜血管分割算法普遍存在的微小血管细节丢失和病灶信息误判等问题,提出一种基于改进HRNet的血管分割算法.在预处理阶段,利用限制对比度自适应直方图均衡化和自适应的Gamma矫正提高血管与背景对比度;在编码阶段,将HRNet原始卷积替换为可变形卷积,提升卷积对复杂血管形态结构的适应能力;在多尺度特征融合阶段,引入空间金字塔池化和多尺度卷积,扩大感受野同时增强对目标局部特征关注度,改善血管伪影和细微信息丢失的问题.该算法在DRIVE数据库上仿真实验,其准确率、灵敏度和特异性分别为95.79％、80.33％和98.12％.     

融合可变形卷积网络的鱼眼图像目标检测

环视鱼眼图像具有目标形变大和图像失真的缺点,导致传统网络结构在对鱼眼图像进行目标检测时效果不佳。为解决环视鱼眼图像中由于目标几何畸变而导致的目标检测难度大的问题,提出一种基于可变形卷积网络的鱼眼图像目标检测方法。将Cascade_RCNN中固定的卷积层和池化层分别替换为可变形卷积层和可变形池化层,使用Resnet50网络提取候选区域以获得检测框,级联具有不同IoU阈值的检测网络进行检测框抑制。在公开鱼眼图像数据集SFU_VOC_360和本文所采集的真实道路场景鱼眼图像数据集上进行实验,结果表明,该方法在鱼眼图像目标检测中具有有效性,目标检测准确率高于Cascade_RCNN网络。     

Fuzzy deduction methods of selecting the underwater terrain navigation matching area

To overcome the problems of not describing terrain matching characteristics totally with a single terrain feature parameter in underwater navigation terrain database, an evaluation method named Vague set is proposed to describe the terrain suitability comprehensively based on the fuzzy decision making (FDM) method. The membership functions and non-membership functions of the Vague set are used to define the influence of the terrain map characteristics parameters on matching suitability, and comprehensive evaluation values of matching areas are obtained by the weighted score function method. Simulation results demonstrate that the terrain map suitability is positively proportional to the comprehensive evaluation value, while the matching position error is negatively proportional to the comprehensive evaluation value of the matching area. The Vague set can choose the optimal matching areaeffectively, and solve the problem that the single topographic map featurecannot evaluate suitability comprehensively.