Papers by Author: Qin Zhang

Abstract: In order to capture high quality binocular stereo video, it is necessary to manipulate both the convergence and the interaxial to take control of the depth of objects within the 3D space. Therefore the scene understanding becomes important as it can increase the efficiency of parameters control. In this paper, a camera calibration based multi-objects location method is introduced with the motivation that supply prior information of adjusting the convergence and the interaxial during capturing. Firstly, we are intended to calibrate the two cameras to get the intrinsic and extrinsic parameters. And then, we select points of the object in the images taken by left and right cameras respectively to determine its locations in the two images. With three-dimensional coordinate of objects, the distance between the object and camera baseline is calculated by mathematical methods.

Abstract: This paper is focused on camera calibration, image matching, both of which are the key issues in three-dimensional (3D) reconstruction. In terms of camera calibration firstly, we adopt the method based on the method proposed by Zhengyou Zhang. In addition to this, it is selective for us to deal with tangential distortion. In respect of image matching, we use the SIFT algorithm, which is invariant to image translation, scaling, rotation, and partially invariant to illumination changes and to affine or 3D projections. It performs well in the follow-up matching the corresponding points. Lastly, we perform 3D reconstruction of the surface of the target object. A Graphical User Interface is designed to help us to realize the key function of binocular stereo vision, with better visualization. Apparently, the entire GUI brings convenience to the follow-up work.

Abstract: In block transform based coding, coefficients scanning plays an important role in improving compression performance. A good scanning method has great potential to remove statistical redundancy, as it can apply an energy compaction coefficients vector. In this paper, a mode-dependent (MD) coefficients scanning method is introduced, with the idea of reducing the overall bit rate by adopting side information to coding mode indexes. In the experiment, we design 128 MD scanning tables according to the DCT coefficients energy distribution, and explore the relationship between the number of scanning table used and the bit saving in image coding. The results provide reference to choose scanning modes by considering the trade-off of the increasing of side information and the decrease of data redundancy.

Abstract: Intra prediction is a key step in H.264/AVC to improve the coding performance with the idea that removing the directional redundancy among neighboring blocks. In order to cover more directional information existed in the image frames, there are usually many prediction modes can be selected in the state-of-the-art coding frameworks, but more bits are also needed to encode the prediction mode index information, then how to achieve the maximum overall bit-rate reduction became a problem. In this paper, 16 kinds of prediction modes are adopted by considering the direction information for 8x8 image blocks. Through calculating the bit-rate both for the mode index and residual image under different number of prediction modes, we obtain the most suitable prediction mode number relatively from the graphs. Experimental results show that, with the increase of prediction mode number, the residual information decreases obviously, and the sum of residual information and prediction mode index information also decreases but levels off after reaching a certain mode number, even has an obviously rising trend.

Abstract: This paper aims at the difficulty that lack of observation model and high-dimensional sampling in video tooning, proposes a method based on key frame matching and dual-directional Markov chain Monte Carlo sampling of video motion redirection. At first, after extracting the key frame of a given video, By affine transformation and linear superposition, the subject initializes the video’s space-time parameters and forms the observation model; Secondly, in each space-time, based on the bi-directional Markov property of each frame, This paper proposed a dual-directional Markov chain Monte Carlo sampling particle filter structure and takes full advantage of the relationship of the front and back frame of the parameters to estimate motion redirection parameters. At the same time, for high-dimensional sampling problem, the subject according to the directional parameters’ correlation implements classification of skeleton parameters-morphological parameters-physical parameters, proposes a hierarchical genetic strategy to optimize the output parameters and improves the efficiency of the algorithm. The research of this paper will produce an efficient and prominent animation expressive video motion redirection method and play an important role on video animation of the development.

Abstract: This paper analyses the motion of animation with the methods of picking key points and predicting motion trace based on the particle filter for discrete particle track prediction theory. The behavior model has been built for the already existing animation character. During the research, the thesis realized using existed animation motion trace model to drive a similar figure and create a new animation.

Abstract: In the process of particle filter, it can produce the problem of particle degeneration, and the amount of computation is very big. In order to solve these two problems, this paper presents a method for video redirection parameters optimization based on hierarchical genetic optimization strategy. This paper adopts hierarchical mutation operator and local crossover operator. The result shows that the hierarchical genetic optimization algorithm which is used in video tracking has a very good effect.

Abstract: Loudspeaker is a strongly nonlinear system which is associated with several situations such as electronic, magnetic, mechanical and acoustic. Recently, most of the methods used to measure and analysis loudspeaker are based on the FFT. Unfortunately, traditional Fourier transform based signal analysis method usually causes meaningless results when it is used to analysis non-stationary and nonlinear signal. In this paper, we use Hilbert-Huang transform (HHT) to review the instantaneous frequency of loudspeaker output. Experiments demonstrate that the distortion of loudspeaker can be recognized as intrawave frequency modulation caused by wave profile deformation. Then a novel nonlinear distortion measurement method is proposed which can reveal more accurate and physical meaningful characteristic of loudspeaker.

Abstract: This paper is based on the particle filter for discrete particle track prediction theory, analyses the motion of animation with the methods of picking key points and predicting motion trace by utilizing particle filter. The behavior model has been built for the already existing animation character. During the research, the thesis realized using existed animation motion trace model to drive a similar figure and create a new animation.

Abstract: SiC/C (graphite) composite ceramics with 20vol% flake graphite were fabricated by Plasma Activated Sintering (PAS) firstly; the composites treated by silicon infiltration in surface were oxidized from 800oC to 1500oC. Effect of silicon infiltration on the oxidation resistance of SiC/C (graphite) composites was characterized by TG/DTA, SEM and XRD. The results show that the graphite in surface of composites reacted with silicon to form a dense SiC film about 30μm-in-thickness. The oxidation resistance temperature was increased from 600oC to 1300oC. Below 1300oC, the surface of composites by silicon filtration was oxidized to form a dense continuous oxide film to hider the oxidation of graphite in matrix. At 1500oC, the surface oxide film was heavy damaged, and many micropores were produced; inner graphite was oxidized above 1300oC again.