FJFeng Jiang

Feng Jiang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Feng Jiang has a total of 20 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Locality-sensitive competitive matrix regression for image classification

    by Hefeng Yin, Qiongdan Lou, Feng Jiang

    Over the past few years, nuclear norm-based matrix regression (NMR) has achieved superior performance in the field of image recognition. By employing the nuclear norm as the loss metric to preserve the two-dimensional structural information of error images, NMR exhibits strong robustness against structural noise. However, NMR fails to effectively exploit the category information of training data and ignores the local structural relationships among samples. To overcome the above limitations, we design a new classification method named locality-sensitive competitive matrix regression (LSCMR). First, a competitive representation term is introduced into the objective function of NMR, enforcing training samples from distinct categories to competitively characterize the test sample. To further boost the discriminative ability of representation coefficients, LSCMR incorporates the local structural information between samples, which encourages samples of the correct class to contribute more to the representation process. The alternating direction method of multipliers (ADMM) is utilized to solve the objective function of LSCMR. Experimental results on several public datasets verify the superiority of LSCMR. Specifically, on subset 4 and subset 5 of the Extended Yale B dataset, LSCMR achieves recognition accuracy of 97.37% and 62.18%, respectively, which are 9.21% and 20.02% higher than those of NMR. The source code of LSCMR is available at https://github.com/yinhefeng/LSCMR.

  • A Weighted Competitive Nonnegative Representation Method for Image Classification

    by Hefeng Yin, Qiongdan Lou, Xiaoning Song, Feng Jiang

    We present a weighted competitive nonnegative representation (WCNR) method. Specifically, WCNR introduces a competitive term, which leverages the competitive ability of training data from distinct classes to strengthen the connection between the representation and classification phases. In addition, WCNR incorporates a weight constraint. The weight constraint of each class imposed on the representation coefficients endows similar classes with more representation contributions, which boosts the discriminative power of the representation coefficients. To assess the classification performance of WCNR, extensive experiments are carried out on benchmarking datasets. Experimental results confirm that WCNR exceeds the state-of-the-art representation-based classification methods (RBCM) and also surpasses several deep learning approaches. The MATLAB code of WCNR is available at https://github.com/yinhefeng/WCNR .

  • Research on gear defect detection using electromagnetic induction eddy current and infrared thermography

    by Feng Jiang, Yanhong Zhang, Lei Zhao, Hongyue Liu, Zhiqiang Song, Hefeng Yin, Yilin Zhu


  • Open Access

    Modeling and investigation of crack evaluation in metal tube based on magnetic field analysis method

    by Feng Jiang, Shu Lin Liu

    In this paper, an analytical model of eddy current testing of a metal tube has been proposed to evaluate circumferential cracks from the signals of the magnetic field. First, based on Maxwell’s equations in time-harmonic form, magnetic vector potential equations for each region of space are built under two different configurations of the bobbin probe and the encircling probe respectively. Next, the magnetic vector potentials are solved by imposing conditions at the interface. Then, series expressions of the radical and axial magnetic induction intensities have been obtained. In addition, based on the proposed modeling, the influences of cracks at different locations on the surface magnetic field have been analyzed. The results have shown that the radial magnetic induction intensity increases while the axial magnetic induction intensity decreases for the internal and external cracks with the same shape and size on the surface of the metal tube. Moreover, the magnetic induction intensity has shown much more sensitive property to external cracks than it has done to internal cracks for the encircling probe. However, for the bobbin probe, the magnetic induction intensity has demonstrated almost the same sensitivity in detection for both internal and external cracks in the metal tube. The researching results have important theoretical and practical significance for understanding the physical process of eddy current testing, optimizing parameters of the experimental system and building the inversion model.


  • Quantitative Research on Cracks in Pipe Based on Magnetic Field Response Method of Eddy Current Testing

    by Feng Jiang, Shulin Liu, Li Tao

    The quantitative evaluation of defects in eddy current testing is of great significance. Impedance analysis, as a traditional method, is adopted to determine defects in the conductor, however, it is not able to depict the shape, size and location of defects quantitatively. In order to obtain more obvious characteristic quantities and improve the ability of eddy current testing to detect defects, the study of cracks in metal pipes is carried out by utilizing the analysis method of three-dimensional magnetic field in present paper. The magnetic field components in the space near the crack are calculated numerically by using finite element analysis. The simulation results confirm that the monitoring of the crack change can be achieved by measuring the magnetic field at the arrangement positions. Besides, the quantitative relationships between the shape, length of the crack and the magnetic field components around the metal pipe are obtained. The results show that the axial and radial magnetic induction intensities are affected more significantly by the cross-section area of the crack. Bz demonstrates obvious advantages in analyzing quantitatively crack circumference length. Therefore, the response signal in the three-dimensional direction of the magnetic field gets to intuitively reflect the change of the defect parameter, which proves the effectiveness and practicability of this method.


  • Evaluation of cracks with different hidden depths and shapes using surface magnetic field measurements based on semi-analytical modelling

    by Feng Jiang, Shulin Liu

    In this paper, we present a feasibility study for detecting cracks with different hidden depths and shapes using information contained in the magnetic field excited by a rectangular coil with a rectangular cross section. First, we solve for the eigenvalues and the unknown coefficients of the magnetic vector potential by imposing artificial and natural boundary conditions. Thus, a semi-analytical solution for the magnetic field distribution around the surface of a conducting plate that contains a long hidden crack is formulated. Next, based on the proposed modelling, the influences of the different hidden depth cracks on the surface magnetic field are analysed. The results show that the horizontal and vertical components of the magnetic field near the crack are becoming weaker and that the phase information of the magnetic field can be used to qualitatively determine the hidden depth of the crack. In addition, the model is optimised to improve its accuracy in classifying crack types. The relationship between signal features and crack shapes is subsequently established. The modified model is validated by using finite element simulations, visually indicating the change in the magnetic field near the crack.


  • Vision-Based Measurement System for Instantaneous Rotational Speed Monitoring Using Linearly Varying-Density Fringe Pattern

    by Jianfeng Zhong, Shuncong Zhong, Qiukun Zhang, Zhike Peng, Shulin Liu, Yingjie Yu

    The instantaneous rotational speed (IRS) of a rotating machine is key information to understand the machine operation and diagnose the potential faults. A novel nonprojection fringe vision-based system was proposed to realize the measurement of IRS, which was simply composed of an artificial linearly varying-density fringe pattern (LVD-FP) as a sensor and a high-speed camera as a detector. The designed LVD-FP, printed by a normal printer, was pasted around the surface of the rotary shaft completely along the fringe density change direction. Part of the LVD-FP was captured by the high-speed camera. The fringe period density of each imaged LVD-FP intensity changed due to the rotation of the shaft, from which the rotational angle could be obtained. Subsequently, both the instantaneous angular speed (IAS) between two adjacent frames and the IRS could be accurately and efficiently obtained. The influences of the key parameters of the proposed system on the measurement accuracy were investigated by simulations and experiments. Both the simulation and experimental results demonstrated that the proposed system can accurately measure the IAS and correctly track the wide range IRS with the advantage of noncontact and cost effective. In addition, the rotation direction of the shaft could be directly obtained from the sign of the IRS curves. Therefore, the proposed system could provide a novel vision-based approach for IRS measurement, especially when the conventional speed sensor cannot be installed on the rotary shaft. The good performance of the system in IRS measurement makes it attractive for fault diagnosis of rotating machinery.