Blind restoration of real turbulence-degraded image with complicated backgrounds using anisotropic regularization

This paper proposes a novel blind image restoration method based on estimating the point-spread functions by using two real turbulence-degraded images as input. The non-negative constraint and the spatial correlation are transformed mathematically into the penalty terms and added to the objective function. An anisotropic and nonlinear regularization function is proposed to adequately punish the differences of the point spread functions (PSFs) in the process of optimization estimation. Some definitions of weighted second-order differences are given and a fast method to construct the matrix of second-order weighted gradient operator is derived. The PSF values can be quickly estimated. With the estimated PSFs, the true images can be recovered by non-blind restoration methods. Experiment results for the restoration of real turbulence-degraded images with complicated backgrounds support the effectiveness of this proposed method.     

Portable data‐parallel surface reconstruction on a uniform rectilinear grid

With the increasing heterogeneity and on‐node parallelism of high‐performance computing hardware, a major challenge is to develop portable and efficient algorithms and software. In this work, we present our implementation of a portable code to perform surface reconstruction using NVIDIA's Thrust library. Surface reconstruction is a technique commonly used in volume tracking methods for simulations of multimaterial flow with interfaces. We have designed a 3D mesh data structure that is easily mapped to the 1D vectors used by Thrust and at the same time is simple to use and uses familiar data structure terminology (such as cells, faces, vertices, and edges). With this new data structure in place, we have implemented a piecewise linear interface reconstruction algorithm in 3 dimensions that effectively exploits the symmetry present in a uniform rectilinear computational cell. Finally, we report performance results, which show that a single implementation of these algorithms can be compiled to multiple backends (specifically, multi‐core CPUs, NVIDIA GPUs, and Intel Xeon Phi processors), making efficient use of the available parallelism on each. We also compare performance of our implementation to a legacy FORTRAN implementation in Message Passing Interface (MPI) and show performance parity on single and multi‐core CPU and achieved good parallel speed‐ups on GPU. Our research demonstrates the advantage of performance portability of the underlying data‐parallel programming model.     

三维激光扫描系统中曲面空洞的识别与修复

为了解决三维激光扫描系统中重构曲面存在的空洞问题,提出了基于Floyd最短路径选择算法的空洞识别与修复方法。该方法对三维曲面中所有可能构成空洞的边界点进行逐个处理,采用树搜索算法获得与处理点直接或间接相连的边界点;将搜索到的边界点作为路径选择的节点,将连接节点的边界边作为路径选择的边并根据节点的搜索级别设置边的长度。当新搜索到的边界点与已搜索点发生重复时,首先,利用Floyd算法处理距离矩阵和路由矩阵找到空洞端点;然后,根据重复点与空洞端点生成空洞边集,最后,采用波前法对空洞边集进行处理。实验结果表明:本文所提方法能够准确识别连接有孤立边的空洞以及两个相邻空洞的特殊空洞结构,与传统方法相比,该方法具有更强的通用性和鲁棒性,空洞修复数量与两个传统方法相比分别提高了54.1%和21.3%。     

Combining multiset resolution and segmentation for hyperspectral image analysis of biological tissues

Hyperspectral images can provide useful biochemical information about tissue samples. Often, Fourier transform infrared (FTIR) images have been used to distinguish different tissue elements and changes caused by pathological causes. The spectral variation between tissue types and pathological states is very small and multivariate analysis methods are required to describe adequately these subtle changes.     

A new volume‐preserving and continuous interface reconstruction method for 2D multi‐material flow

A new two‐dimensional interface reconstruction method that ensures continuity of the interface and preserves volume fractions is presented here. It is made of two steps, first, the minimization of a cost functional based on volume fractions least square errors by using dynamic programming, a fast and efficient scheme well known in image processing, and then a local correction phase. In each cell, the interface is made of two line segments joining two edges. This new interface reconstruction method, called Dynamic Programming Interface Reconstruction has been coupled with various advection schemes, among them the Lagrange + remap scheme. With a reasonable computational cost, it has been observed in various test cases that Dynamic Programming Interface Reconstruction is more accurate and less diffusive compared with other existing classical reconstruction methods. Copyright © 2017 John Wiley & Sons, Ltd.     

A simple quality evaluation method of binary images based on Border Distance

An objective quality evaluation method of binary images is proposed. A new distance measurement, Border Distance, is designed to show the location effects of modified pixel. The quality evaluation of binary image based on Border Distance considers the image resolution and the Border Distance between the modified pixels and the border. Experimental results show that the proposed method well matches the human visual perception.     

Implicit large eddy simulation using the high‐order correction procedure via reconstruction scheme

We investigate implicit large eddy simulation of the Taylor–Green vortex, Comte‐Bellot–Corrsin experiment, turbulent channel flow and transitional and turbulent flow over an SD7003 airfoil using the high‐order unstructured correction procedure via reconstruction (CPR) scheme, also known as the flux reconstruction scheme. We employ P1 (second‐order) to P5 (sixth‐order) spatial discretizations. Results show that the CPR scheme can accurately predict turbulent flows without the addition of a sub‐grid scale model. Numerical dissipation, concentrated at the smallest resolved scales, is found to filter high‐frequency content from the solution. In addition, the high‐order schemes are found to be more accurate than the low‐order schemes on a per degree of freedom basis for the canonical test cases we consider. These results motivate the further investigation and use of the CPR scheme for simulating turbulent flows. Copyright © 2016 John Wiley & Sons, Ltd.     

Analysis of the correlation between plasma plume and keyhole behavior in laser metal welding for the modeling of the keyhole geometry

The process of laser metal welding is widely used in industry. Nevertheless, there is still a lack of complete process understanding and control. For analyzing the process we used two high-speed cameras. Therefore, we could image the plasma plume (which is directly accessible by a camera) and the keyhole (where most of the process instabilities occur) during laser welding isochronously. Applying different image processing steps we were able to find a correlation between those two process characteristics. Additionally we imaged the plasma plume from two directions and were able to calculate a volume with respect to the vaporized material the plasma plume carries. Due to these correlations we are able to conclude the keyhole stability from imaging the plasma plume and vice versa. We used the found correlation between the keyhole behavior and the plasma plume to explain the effect of changing laser power and feed rate on the keyhole geometry. Furthermore, we tried to outline the phenomena which have the biggest effect on the keyhole geometry during changes of feed rate and laser power.     

基于大气多散射模型和超像素分割的图像去雾

通过分析大气粒子对光线的多次散射作用,利用大气点扩散函数作为卷积核,基于暗原色图像复原理论,建立了基于多次散射的雾天成像模型,并以函数形状相似性为依据,利用广义高斯分布定量估计出大气点扩散函数核函数在图像域下的相关参数.针对传统暗原色理论以固定大小图像区域估计透射率的不足,提出了基于超像素分割获得景深一致的图像分块方案,通过区域合并,获得更为精准的天空检测效果;基于暗原色先验理论分别估计天空和非天空区域的透射率,并对天空区域的透射率进行修正,不但减少了天空色彩失真,同时也消除了复原结果的光晕现象.本文从主观和客观两个方面将所提出的去雾方法和其他算法进行了对比,结果表明,本文提出的去雾算法能够在较短的运行时间内获得对比度较高、细节信息丰富的去雾结果,具有较好的鲁棒性.