Xingjie Li

Shenyang Research Institute of Foundry

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Han Yu

Shenyang Research Institute of Foundry

Xinyue Li

Shenyang Research Institute of Foundry

Erfeng Shang

Shenyang Research Institute of Foundry

John Hall

Lee College

James Roetzer

Lee College

Yunhui Yan

Northeastern University

Huasheng Xie

Shenyang Research Institute of Foundry

Kechen Song

Macao Polytechnic University

Chunyu Hou

Henan University

Ben Janke

University of North Carolina at Charlotte

Cooperative Institutions

Shenyang Research Institute of Foundry

Fudan University

Northeastern University

University of North Carolina at Charlotte

Shanghai University of Finance and Economics

Chinese Academy of Sciences

University of Chinese Academy of Sciences

Ministry of Education of the People's Republic of China

Sichuan University

State Key Laboratory of Light Alloy Casting Technology for High-end Equipment

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Improving Scattered Defect Grading in Castings Digital Radiographs via Smoothing the One-Hot Encoding

International Journal of Metalcasting (2024)

Han Yu Xingjie Li Hao Xue Zhaowei Song Shangyu Liu

Structural material

Smoothing

Grading (engineering)

10.1007/s40962-024-01335-3

Cite

Citations (0)

3D characterisation of metal powders for additive manufacturing using Nano-CT and deep learning

Nondestructive Testing And Evaluation (2024)

Hao Xue Xinyue Li Zhida Jin Xingjie Li Yabin Zhang

The geometrical properties of metal powders, the main raw material in additive manufacturing processes, can have a significant impact on the quality of the parts produced. Therefore, it is essential to characterise these powders effectively. Nano-computed tomography (Nano-CT) technology, with its high-resolution 3D measurement capability, is ideally suited to measure the geometrical characteristics of tiny powders. However, the current CT testing methods for metal powders has two main drawbacks. Firstly, it does not take into account the effect of particle adhesion. Second, it cannot achieve high-precision measurements of the internal pores of powders. To address these limitations, we propose a novel testing method for metal powders that utilises Nano-CT technology. The method first obtains highly dispersed powder samples through a specially sample preparation technique. Then, an improved U-shape networks with multi-level supervision are proposed to achieve high-precision segmentation of the internal pores of the powder. Ultimately, this paper achieves high-precision 3D characterisation of metal powders.

10.1080/10589759.2024.2431146

Cite

Citations (0)

A Distribution-Preserving Under-Sampling Method for Imbalance Defect Recognition in Castings

Coatings (2022)

Yu Han Xinyue Li Xingjie Li Chunyu Hou Shangyu Liu

Data imbalance is a crucial factor that limits the performance of automatic defect recognition systems in castings. The bias and deterioration of the model are generated by massive normal samples and minor defect samples. Traditional re-sampling methods randomly change the data distribution and ignore the significant intra-class difference among all normal samples. Therefore, this paper proposes a distribution-preserving under-sampling method for imbalance defect-recognition in castings. In detail, our method divides all normal samples into several sub-groups by cluster analysis and reassembles them into some balance datasets, which makes the normal samples in all balance datasets have an identical distribution with the original imbalance dataset. Finally, experiments on our dataset with 3260 images indicate that the proposed method achieves a 0.816 AUC (area under curve) score, which demonstrates significant advantages compared to cost-sensitive learning and re-sampling methods.

10.3390/coatings12121808

Cite

Citations (1)

Exploration of SVD for Image Compression and Time Series Processing

Jesse Redford Xingjie Li

This paper explores the block-based singular-value-decomposition (SVD) with applications to image compression and processing in both static and dynamic cases. Results include a comprehensive performance study of rank-1 asymmetric block sizes in comparison to conventional uniform square blocks for static image compression. In addition, a method of movement detection for video streaming applications are presented and discussed in this paper. We show that by tracking the Frobenius norm of the derivatives of sequential images, represented as a time series matrix, allows one to infer change points in varying lighting conditions and foreground movement in video sequences. We show that by combining SVD background subtraction we can create clearer distinction between movements and changes in lighting intensity. It is also demonstrated that one can maintain the original derivative feature after re-scaling the images of the video stream to a fraction of their original resolution via local averaging.

Matrix norm

10.1109/urtc51696.2020.9668891

Cite

Citations (2)

Image Dilution using Harris Corner Detection and Geometric Kernels

Aiden James Xingjie Li

Image files contain large amounts of data. An image is essentially a matrix of values, often represented as ARGB (alpha, red, green, blue) arrays. A large three dimensional matrix with a height and width likely in the hundreds can take long for a program to read. However, many applications may only require specific key features to understand an image. The majority of pixel data is relatively unimportant when determining the contents of the image file. In fact, such extra data can sometimes deceive the machine, or in the case of a hybrid image [1], the human viewer. Using only basic matrix calculations and matrix convolution, specifically Harris Corner Detection and Edge Detection kernels, we have extracted the key features of an image in order to dilute its data. The results show that any confusion, such as noise or the low frequencies of a hybrid image, becomes weaker, and that the computation is very efficient and robust.

Convolution (computer science)

Matrix (chemical analysis)

Corner detection

10.1109/urtc56832.2022.10002246

Cite

Citations (0)

Improving the Industrial Defect Recognition in Radiographic Testing by Pre-training on Medical Radiographs

NDT & E International (2024)

Han Yu Xingjie Li Huasheng Xie Xinyue Li Chunyu Hou

Radiographic testing

10.1016/j.ndteint.2024.103260

Cite

Citations (1)

A Feature Oriented Framework and Enhanced Assessments for Imaging Compression

Jesse Redford Xingjie Li

This paper offers a new feature-oriented compression algorithm for flexible reduction of data redundancy commonly found in images and videos streams. Using a combination of image segmentation and face detection techniques as a preprocessing step, we derive a compression framework to adaptively treat `feature' and `ground' while balancing the total compression and quality of `feature' regions. We demonstrate the utility of a feature compliant compression algorithm (FC-SVD), a revised peak signal-to-noise ratio PSNR assessment, and a relative quality ratio to control artificial distortion. The goal of this investigation is to provide new contributions to image and video processing research via multi-scale resolution and the block-based adaptive singular value decomposition.

Feature (linguistics)

Peak signal-to-noise ratio

10.1109/ipas50080.2020.9334966

Cite

Citations (3)

A Study of Illumination Filtering and Movement Detection in Video Streaming Processing

Mai Phuong Nguyen Xingjie Li

We study a movement detection scheme for video streaming data using time series analysis. We employ simple scalar quantities such as (i) the Frobenius norm trajectory of a forward difference matrix and (ii) the Auto-correlation function to analyze the time series and detect movements in the video sequences. To remove the background interruptions, such as the change of lights, we study the performance of two filters, (1) SVD filter and (2) SVD combined with the Butterworth filter, to repress data disruptions. From the example, we identify the best filter and the best statistical quantity to detect movement in the disrupted data. By using the background subtraction filters as well as the Frobenius norm trajectory of data, we can distinguish the movements and light disruptions at various time points in the video sequences. The combo filter is found to be the best as it completely removes the light interruption while maintaining the movements data unaffected and smooths other static artifacts.

Matrix norm

Subtraction

10.1109/urtc54388.2021.9701635

Cite

Citations (0)

Data-Driven Modeling Adaptive Aerostructures

James Roetzer Hunter Boik Ben Janke Xingjie Li John Hall

Abstract In the growing domain of wind energy, morphing aerostructures provide an avenue of research to improve aerodynamic performance and turbine control. Branching off previous work by the authors, several data-driven models which relate the actuation of a linearly segmented active morphing blade to an arbitrary performance parameter, aerodynamic thrust load in the case examined, are developed, presented, and compared to evaluate accuracy. Models are developed for the Danish Technical University (DTU) 10 MW and National Renewable Laboratory (NERL) 15 MW reference turbines to ensure generalizability. The models are trained using OpenFAST simulations to generate training and test data. State vector machine (SVM), neural network (NN), decision tree, and gaussian process regression (GPR) models are developed. Comparing root-mean-square error (RMSE), the GPR method produces the least error, in most cases by an order of magnitude or more. The developed model is then used in a demonstrative application, an optimization of the target parameter. Particle swarm optimization is used to find the global minimum input space of the GPR model at discrete wind speeds. The references turbines show a respective maximum improvement in the target performance parameter of 3.15% and 2.66%. Optimization results indicate that refinement of the developed model may improve results.

10.1115/detc2024-145979

Cite

Citations (0)

COVID-19 in Japan: What could happen in the future?

medRxiv (Cold Spring Harbor Laboratory) (2020)

Nian Shao Yan Xuan Hanshuang Pan Shufen Wang Weijia Li

COVID-19 has been impacting on the whole world critically and constantly Since December 2019. We have independently developed a novel statistical time delay dynamic model on the basis of the distribution models from CCDC. Based only on the numbers of confirmed cases in different regions in China, the model can clearly reveal that the containment of the epidemic highly depends on early and effective isolation. We apply the model on the epidemic in Japan and conclude that there could be a rapid outbreak in Japan if no effective quarantine measures are carried out immediately.

Isolation

2019-20 coronavirus outbreak

Epidemic model

Pandemic

Containment (computer programming)

10.1101/2020.02.21.20026070

Cite

Citations (8)