Mingwei Sheng

Prof. Mingwei Sheng

orcid

Research Focus & Metrics

Academic Career Length

16Years (2011-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 February 2021 in Robotica
Verified
Journal Article
Closed access
PDF file
CSChaoyu Sun
Chaoyu Sun
ZWZhaoliang Wan
Zhaoliang Wan
HHHai Huang
Hai Huang
GZGuocheng Zhang
Guocheng Zhang
XBXuan Bao
Xuan Bao
Xu YangXu Yang
Xu Yang
SUMMARY: Visual tracking is an essential building block for target tracking and capture of the underwater vehicles. On the basis of remotely autonomous control architecture, this paper has proposed an improved kernelized correlation filter (KCF) tracker and a novel fuzzy controller. The model is trained to learn an online correlation filter from a plenty of positive and negative training samples. In order to overcome the influence from occlusion, the improved KCF tracker has been designed with an added self-discrimination mechanism based on system confidence uncertainty. The novel fuzzy logic tracking controller can automatically generate and optimize fuzzy rules. Through Q-learning algorithm, the fuzzy rules are acquired through the estimating value of each state action pairs. An S surface based fitness function has been designed for the improvement of learning based particle swarm optimization. Tank and channel experiments have been carried out to verify the proposed tracker and controller through pipe tracking and target grasp on the basis of designed open frame underwater vehicle.

1 March 2020 in International Journal of Advanced Robotic Systems
Verified
Journal Article
Open access
PDF file
MSMingwei Sheng
Mingwei Sheng
STSongqi Tang
Songqi Tang
ZCZhuang Cui
Zhuang Cui
WWWanqi Wu
Wanqi Wu
LWLei Wan
Lei Wan
Panoramic stitching technology provides an effective solution for expanding visual detection range of the autonomous underwater vehicle. However, absorption and scattering of light in the water seriously deteriorate the underwater imaging in terms of distance and quality, especially the scattering sharply decreases the underwater image contrast and results in serious blur. This reduces the number of matching feature points between the underwater images to be stitched, while fewer matched points generated make image registration and stitching difficult. To solve the problem, a joint framework is established, which firstly involves a convolutional neural network-like algorithm composed of a symmetric convolution and deconvolution framework for underwater image enhancement. Then, it proposes an improved convolutional neural network-random sample consensus method based on VGGNet-16 framework to generate more correct matching feature points for image registration. The fusion method based on Laplacian pyramid is applied to eliminate artificial stitching traces and correct the position of stitching seam. Experimental results indicate that the proposed framework can restore the color and detail information of underwater images and generate more effective and sufficient matching feature points for underwater sequence images stitching.

1 May 2020 in International Journal of Advanced Robotic Systems
Verified
Journal Article
Open access
PDF file
LWLei Wan
Lei Wan
GCGuofang Chen
Guofang Chen
MSMingwei Sheng
Mingwei Sheng
YZYinghao Zhang
Yinghao Zhang
ZZZiyang Zhang
Ziyang Zhang
This study investigates an adaptive chattering-free sliding-mode control method for n-order nonlinear systems with unknown external disturbances and uncertain models. The proposed method takes the advantage of finite-time fast convergence to avoid singularity problem and ensure its robustness against system uncertainty and unknown disturbance. To achieve fast convergence from any initial condition to system origin, a full-order terminal sliding-mode controller containing differential terms is proposed based on the property of n-order nonlinear systems. Then the continuous and smooth actual control law is obtained by integrating the differential control law containing the discontinuous sign function to realize chattering free. Meanwhile, instead of evaluating the fixed upper bound of system uncertainty and interference in practical implementations, an adaptive method is utilized for its unknown upper bound estimation. The convergence of the adaptive terminal sliding-mode controller in finite time is verified based on Lyapunov stability theory. Finally, two simulation results demonstrate the effectiveness of the proposed control method.

17 January 2019 in Sensors
Verified
Journal Article
Open access
PDF file
MSMingwei Sheng
Mingwei Sheng
STSongqi Tang
Songqi Tang
HQHongde Qin
Hongde Qin
LWLei Wan
Lei Wan
Autonomous underwater vehicles (AUVs) rely on a mechanically scanned imaging sonar that is fixedly mounted on AUVs for underwater target barrier-avoiding and tracking. When underwater targets cross or approach each other, AUVs sometimes fail to track, or follow the wrong target because of the incorrect association of the multi-target. Therefore, a tracking method adopting the cloud-like model data association algorithm is presented in order to track underwater multiple targets. The clustering cloud-like model (CCM) not only combines the fuzziness and randomness of the qualitative concept, but also achieves the conversion of the quantitative values. Additionally, the nearest neighbor algorithm is also involved in finding the cluster center paired to each target trajectory, and the hardware architecture of AUVs is proposed. A sea trial adopting a mechanically scanned imaging sonar fixedly mounted on an AUV is carried out in order to verify the effectiveness of the proposed algorithm. Experiment results demonstrate that compared with the joint probabilistic data association (JPDA) and near neighbor data association (NNDA) algorithms, the new algorithm has the characteristic of more accurate clustering.

24 February 2020 in Electronics
Verified
Journal Article
Open access
PDF file
MSMingwei Sheng
Mingwei Sheng
WWWeizhe Wang
Weizhe Wang
HQHongde Qin
Hongde Qin
LWLei Wan
Lei Wan
JLJun Li
Jun Li
WWWeilin Wan
Weilin Wan
Athlete detection in sports videos is a challenging task due to the dynamic and cluttered background. Distractor-aware SiamRPN (DaSiamRPN) has a simple network structure and can be utilized to perform long-term tracking of large data sets. However, similarly to the Siamese network, the tracking results heavily rely on the given position in the initial frame. Hence, there is a lack of solutions for some complex tracking scenarios, such as running and changing of bodies of athletes, especially in the stage from squatting to standing to running. The Haar feature-based cascade classifier is involved to catch the key frame, representing the video frame of the most dramatic changes of the athletes. DaSiamRPN is implemented as the tracking method. In each frame after the key frame, a detection window is given based on the bounding box generated by the DaSiamRPN tracker. In the new detection window, a fusion method (HOG-SVM) combining features of Histograms of Oriented Gradients (HOG) and a linear Support-Vector Machine (SVM) is proposed for detecting the athlete, and the tracking results are updated in real-time by fusing the tracking results of DaSiamRPN and HOG-SVM. Our proposed method has reached a stable and accurate tracking effect in testing on men’s 100 m video sequences and has realized real-time operation.

31 October 2011 in International Journal of Mechanical Sciences
Verified
Journal Article
Closed access
MSMingwei Sheng
Mingwei Sheng
WCWeishan Chen
Weishan Chen
JLJunkao Liu
Junkao Liu
SSShengjun Shi
Shengjun Shi