一种改进的各向异性扩散深度图像增强算法

深度图像受其测距原理所限,存在边缘不匹配、无效像素、噪声等问题,提出一种基于改进的各向异性扩散算法的深度图像增强方法。首先,校正深度图像和彩色图像的位置关系,并根据时间连续性选择多帧图像,进行多帧均值滤波预处理;其次,通过在彩色图像中引入权重的思想,构建具有4-邻域形式的深度图像模型,利用彩色图像引导的深度图像进行各向异性扩散,填补孔洞;最后,使用改进的自适应中值滤波平滑图像噪声。实验结果表明,该方法能够有效修复原始深度图像中存在的由无效像素组成的黑色孔洞,在抑制噪声的同时,仍能保持深度图像中物体边缘的细节信息。     

Performance of estimating depth in projection based stereoscopic virtual display

Accurate distance estimation is essential for effective user interaction with objects appearing in the virtual space. Research has shown that the distance estimation in the virtual space is not as accurate as that in the real world. The inaccuracy phenomenon is almost exclusively documented for estimating objects appearing straight ahead of the eye. A collection of the accuracy data for targets located in different angles of view in the visual field may be useful in aiding the design of user‐virtual object interaction. This study collected the object to object distance estimation accuracy for targets displayed stereoscopically. There are five horizontal and two vertical viewing angles and three depths, resulting in a total of 30 locations in the virtual space. The distance estimation for any of the 30 object pairs were performed with or without the aid of spatial cues. Significant factors influencing the accuracy and time of distance estimation were identified. Implications of the finding were discussed with respect to the stereoscopic display environment.     

考虑桩顶水平反力的水平受荷单桩弹塑性解析解

基于线性地基反力法和土体的弹塑性本构关系,对桩顶水平反力系数不为零时的水平受荷桩进行了研究,给出了桩顶土屈服临界荷载（水平力、弯矩）的耦合关系式。研究结果表明,当荷载大于临界荷载时,桩顶土发生塑性屈服,其塑性屈服深度随荷载的增加而增加,但其塑性区扩展速率随桩顶荷载的增加而逐渐减小。给出了弹性区与塑性区桩内力耦合的矩阵方程及其迭代求解格式;利用曲面拟合软件给出了塑性区换算深的计算公式。通过工程算例验证,证明该方法可靠快捷,便于手算,值得推广。      

海上地震拖缆自动定深器建模及仿真研究

本文首先对自行研制的海上地震拖缆自动定深器建立了较为详细的数学模型,给出了其传递函数模型,然后在Matlab环境下,利用仿真工具Simulink和控制系统工具箱（ControlSystemToolbox）中的SISO工具设计了经典控制器,并且对该系统进行了仿真分析,仿真结果表明：利用本文所设计的控制器,该系统具有较好的动态特性。     

后期模拟制作相机镜头背景模糊效果

在摄影实践中,虚化背景是突出主体的常用手段。但是由于消费级DC镜头的实际焦距都很短,因此实现浅景深而虚化背景的难度较大。如果我们希望用消费级DC也能达到虚化背景突出主体的效果,那么,Photoshop后期处理可以派上用场。     

浅谈随机决策树

该文介绍了随机决策树分类模型及如何启发式选择随机决策树的深度及棵树,通过实验证明了该算法的有效性和高效性。     

具有深度自适应估计的视觉伺服优化

在手眼机器人视觉伺服中,如何确定机器人末端摄像机移动的速度和对物体的深度进行有效的估计还没有较好的解决方法.本文采用一般模型法,通过求解最优化控制问题来设计摄像机的速度,同时,利用物体初始及期望位置的深度估计值,提出了一种自适应估计的算法对物体的深度进行估计,给出了深度变化趋势,实现了基于图像的定位控制.该方法能够使机器人在工作空间范围内从任一初始位置出发到达期望位置,实现了系统的全局渐近稳定且不需要物体的几何模型及深度的精确值.最后给出的仿真实例表明了本方法的有效性.     

对含内孔面模型的切割仿真研究

为了快速封堵含内孔的面模型被切割后形成的中空截面轮廓,以便完整的表达模型.采用先剖分外轮廓,以封堵整个截面的轮廓,然后依次去除内轮廓区域的思想,最终实现中空面的封堵.先使用逐点插入算法剖分整个中空截面轮廓,然后删除已检测到的内孔洞的区域,在确保剖切面内所有轮廓顶点不发生任何变化的要求下.实现了含内孔面模型的快速切割仿真,并使用实例加以验证.该算法可以广泛地应用于各种含内孔面模型的切割仿真,尤其在医学手术仿真中.     

Characterization of laser hardened steels by laser induced ultrasonic surface waves

Ultrasonic surface waves are suitable for the characterization of surface hardened materials. This is shown on laser hardened turbine blades. The martensitic microstructure within the surface layer of surface hardened steels has a lower surface wave propagation velocity than the annealed or normalized substrate material. Because the propagation velocity depends on the ratio of layer thickness to wavelengthd/, its measurement allows the determination of the hardening depth. If the surface wave frequency is high enough, the surface wave propagates mainly within the hardened layer. A correlation of the surface wave velocity to the surface hardness has been found. Because the variation of the surface velocity in hardened steels is small, a high measurement accuracy is necessary to obtain the interesting hardening parameters with sufficient certainty. Therefore, a measuring arrangement has been developed where laser pulses, guided by optical fibers to the surface hardened structure, generate simultaneously surface wave pulses at two different positions. The two ultrasonic pulses are received by a piezoelectric transducer. The surface wave velocity is obtained from the time delay between these pulses which is determined by the cross-correlation method. To evaluate simultaneously surface waves with different penetration depths from the same signal acquisition, digital filtering has been used in connection with the cross-correlation.     

基于改进YOLOv3的智慧足球场行人检测

由于足球比赛场景中密集人群、移动小目标居多, YOLOv3算法存在检测精确度较低且模型参数量较大等问题, 使其无法部署在资源算力有限的移动设备上, 本文提出了一种基于改进YOLOv3的行人检测方法, 将Darknet-53主干特征提取网络替换为更加高效且轻量化的GhostNet网络; 同时选取了4个尺度的检测分支层并采用K-means++算法改善anchor box的聚类效果; 添加空间金字塔池化对输入图像实现相同大小的输出; 提出CIoU损失函数来计算目标定位损失值; 添加heatmap热力图可视化并在训练中使用Mosaic数据增强. 实验结果表明, YOLOv3-GhostNet在VOC融合数据集上mAP达到90.97%的同时相比YOLOv3算法提高了1.75%, 参数量减少了约81.4%且实时检测速率提高了约1.5倍, 在小型移动设备上表现出不错的检测效果.