Beijing Institute of Graphics

Beijing, China

Beijing Institute of Graphics

Beijing, China

Time filter

Source Type

Xiao P.,Beihang University | Duan M.,Beihang University | Zhao Q.,Beijing Institute of Graphics
Beijing Hangkong Hangtian Daxue Xuebao/Journal of Beijing University of Aeronautics and Astronautics | Year: 2012

A novel visual tracking approach based on multiple-kernel fusion was proposed to improve robustness and accuracy of tracking under large-area occlusion. Unlike traditional single symmetric kernel weighted histogram used in mean shift tracking, this approach adopted several asymmetric kernel functions centered at different positions within target region to build a set of asymmetric kernel weighted histograms. Because these histograms weighted each part of the target region differently, there must be some less influenced histograms during occlusion. Based on each histogram, a set of target location estimations were provided respectively by mean shift iteration, and the target location was obtained by fusing these estimations using Dempster-Shafer evidence theory. The experimental results demonstrate the effectiveness of the proposed approach under large-area occlusion.


Hu H.-M.,Beihang University | Zhou M.,Beihang University | Liu Y.,Beijing Institute of Graphics | Yin N.,Beihang University
Multimedia Tools and Applications | Year: 2016

During I frame switching, the subjective quality between I frame and P frames usually have obvious fluctuation due to different coding methods. The periodic temporal visual fluctuation will cause video flicker. According to extensive experiments, we observe that I frame flicker possess a strong regional characteristics and different region have different degree of flicker. Based on this observation, a region-based I frame rate control scheme is proposed to suppress I frame flicker according to the different characteristics of the moving and non-moving regions. Firstly, by jointly considering the inter-frame dependency between I frame and subsequent un-encoded P frames and the inter-frame correlation between I frame and previous encoded P frame, an optimization model is proposed to achieve the optimal QPs for different regions. Secondly, a region-based inter-frame dependency model is proposed to separately describe the inter-frame dependency of different regions, which can accurately describe their description of the inter-frame dependency. The experimental results demonstrate that the proposed scheme can efficiently suppress I frame flicker and maintain the smoothness of subjective quality. Moreover, the proposed scheme can achieve a PSNR gain by 0.26 dB on average when compared with the rate control scheme adopted by the HEVC reference software HM15.0. © 2016 Springer Science+Business Media New York


Zuo B.,Yantai Naval Aeronautical and Astronautical University | Zuo B.,Beihang University | Li J.,Yantai Naval Aeronautical and Astronautical University | Li J.,Beijing Institute of Graphics | Huang H.-L.,Shijiazhuang Army Command College
Xi Tong Gong Cheng Yu Dian Zi Ji Shu/Systems Engineering and Electronics | Year: 2013

A systematic procedure for synthesis of neural network adaptive synergetic control is proposed for a class of affine multivariable extremum seeking system. By employing the synergetic control, the synergetic convergence among the states can be realized, and the invariance against system parameter variation and external perturbation can also be achieved. By using the system's states and intput, the search variables from the extremum seeking control, and the known model parameters as the inputs, two three-layer neural networks are designed to estimate the dynamic process of the states extrema and unknown parameters, respectively. At the same time, an adjustable parameter is used to minify the estimation errors of the three-layer neural networks. The detailed theoretical analysis proves that all errors of the closed-loop system exponentially converge to a small tunable neighborhood of the origin by appropriately choosing design constants. Simulation results show the effectiveness of the proposed control method.


Zuo B.,Yantai Naval Aeronautical and Astronautical University | Li J.,Yantai Naval Aeronautical and Astronautical University | Li J.,Beijing Institute of Graphics
Kongzhi Lilun Yu Yingyong/Control Theory and Applications | Year: 2013

In the proposed synergetic control, the synergetic convergence of states can be realized, and the invariance against the system parameter variation and external perturbation can also be achieved. By using the search variables from the extremum-seeking control as the inputs, multilayer neural networks (MNN) are applied to approximate the differential of the state extrema as well as unknown parameters and functions. The problem of the unknown control gain is well solved by using Nussbaum gain function. At the same time, an adaptive parameter is adopted to compensate for the influence of MNN approximation errors. The stability analysis shows that tracking errors of states, errors between the output and its extrema, tracking errors of search variables, and estimation errors of MNN parameters, all converge exponentially to a small neighborhood of the origin by appropriately choosing design parameters. Theoretical analysis and simulation results show the effectiveness of the proposed control method.


Zhao J.,Xi'an Jiaotong University | Han C.,Xi'an Jiaotong University | Wei B.,Xi'an Jiaotong University | Zhao Q.,Beijing Institute of Graphics | And 2 more authors.
15th International Conference on Information Fusion, FUSION 2012 | Year: 2012

Feature selection is an effective data preprocessing step to reduce the dimension of feature space and save storage space. Binary particle swarm optimization (BPSO) has been applied successfully to solve feature selection problem. But it was easy to fall into local optimal point. M2BPSO was an improved BPSO algorithm. The particles of M2BPSO were updated by using various evolutionary strategies according to the performance of them, in addition, the mutation was used to overcome premature. In this paper, we adopted M2BPSO to solve feature selection problem. To test the validity of the algorithm, we compared it with various versions of BPSO methods. Experimental results showed that M2BPSO could effectively solve the feature selection problem. © 2012 ISIF (Intl Society of Information Fusi).


Li J.,Yantai Naval Aeronautical and Astronautical University | Li J.,Beijing Institute of Graphics | Zuo B.,Yantai Naval Aeronautical and Astronautical University | Duan M.,Beijing Institute of Graphics | Zhang J.,Beijing Institute of Graphics
Hangkong Xuebao/Acta Aeronautica et Astronautica Sinica | Year: 2012

An adaptive terminal sliding mode control method based on the inner-loop and outer-loop systems is designed and analyzed for the longitudinal dynamics of air-breathing hypersonic vehicles under control input constraints. Considering the angle of attack and the pitch rate as the states of the inner-loop system with aeroelastic modes and disturbances, an adaptive terminal sliding mode control law of the elevator deflection is proposed based on backstepping design. Considering the velocity as the state of the outer-loop system, a dynamic inversion adaptive control law of the equivalence ratio is also obtained based on dynamic inversion design. Moreover, two multilayer neural networks are used for the approximation of the saturation property of two control inputs, respectively. Based on Lyapunov stability theorem, all signals of the closed-loop system are bounded and exponentially converge to the neighborhood of the origin globally. Furthermore, aeroelastic modes of air-breathing hypersonic vehicles are asymptotically stable. Simulation results demonstrate the effectiveness of the proposed control method in saturation alleviation of the longitudinal dynamics of air-breathing hypersonic vehicles. In additional, the control method not only realizes the control aim, but also improves transient performance of the controlled system.

Loading Beijing Institute of Graphics collaborators
Loading Beijing Institute of Graphics collaborators