Yanbing Xue

Tianjin University of Technology, T’ien-ching-shih, Tianjin Shi, China

Are you Yanbing Xue?

Claim your profile

Publications (11)1.57 Total impact

  • Zan Gao · Hua Zhang · Anan A. Liu · Guangping Xu · Yanbing Xue
    [Show abstract] [Hide abstract]
    ABSTRACT: Human action recognition is a hot research topic; however, the change in shapes, the high variability of appearances, dynamitic background, potential occlusions in different actions and the image limit of 2D sensor make it more difficult. To solve these problems, we pay more attention to the depth channel and the fusion of different features. Thus, we firstly extract different features for depth image sequence, and then, multi-feature mapping and dictionary learning model (MMDLM) is proposed to deeply discover the relationship between these different features, where two dictionaries and a feature mapping function are simultaneously learned. What is more, these dictionaries can fully characterize the structure information of different features, while the feature mapping function is a regularization term, which can reveal the intrinsic relationship between these two features. Large-scale experiments on two public depth datasets, MSRAction3D and DHA, show that the performances of these different depth features have a big difference, but they are complementary. Further, the features fusion by MMDLM is very efficient and effective on both datasets, which is comparable to the state-of-the-art methods.
    No preview · Article · Sep 2015 · Neural Computing and Applications
  • Qi Dong · Guangping Xu · Liu Jing · Yanbing Xue · Hua Zhang
    [Show abstract] [Hide abstract]
    ABSTRACT: The contour analysis and identification are the important aspects in visual surveillance research. The paper proposes a fuzzy identification method of contours. First, according to the description of a contour based on the chain-code method, the proposed method utilizes the statistical features of contours including the chain-code entropy and chain-code space distribution entropies, from which the feature vector of a contour is composed. Then, the method generates the contour pattern from some contour samples and uses the approaching principle to identify a contour. Since our method integrates effectively multiple statistical feathers of its chain-code and employs a fuzzy pattern recognition technique, the experiments show quantitatively that it can achieve good results from various metrics.
    No preview · Article · Dec 2011

  • No preview · Conference Paper · Jul 2011
  • Qian Wang · Hua Zhang · Qi Dong · Qingxiao Niu · Guangping Xu · Yanbing Xue
    [Show abstract] [Hide abstract]
    ABSTRACT: Since Otsu algorithm does not take the image spatial neighbor information into consideration, we combine the Markov random field with Otsu algorithm to integrate gray level information and spatial correlation information for the pixels. In this paper, Otsu thresholding algorithm based on Markov Random Field is proposed. In this algorithm, the neighborhood rejectability function is imported to Otsu algorithm and an threshold selection function is improved. The experiment results verify that applying our algorithm to road image segmentation can achieve good effects.
    No preview · Conference Paper · Jan 2011
  • Lan Zhang · Hua Zhang · Simiao Zhang · Yanbing Xue
    [Show abstract] [Hide abstract]
    ABSTRACT: We propose a robust approach of producing a high-resolution (HR) image from a sequence of low-resolution (LR), blurred and noisy images. In the proposed approach, we specifically focus on the motion model of Gaussian Pyramid Optical Flow (GPOF) registration which achieves the sub-pixel precision and enables large pixel motions, while keeping the size of the integration window relatively small. In the process of super-resolution reconstruction, our method is based on the use of L1-norm both in the measurement term and the regularization term called Bilateral Total Variation (BTV) as the prior model to penalize high spatial frequency signals and preserve edges. Specially, we introduce the Median “shift and add” idea to initialize the HR image value in the iterative steps for the optimization of the objective function, when the motions between LR frames are pure translations and the blur is space invariant.
    No preview · Conference Paper · Jan 2010
  • Simiao Zhang · Hua Zhang · Lan Zhang · Yanbing Xue
    [Show abstract] [Hide abstract]
    ABSTRACT: We propose a new algorithm about multi-scale-based super-resolution on face image. First, steerable pyramid is used to capture low-level local features in face images, and then these features are combined with pyramid-like parent structure and image patch synthetic approach based on neighborhood to predict the best prior. After that, the prior is integrated into Bayesian maximum a posteriori (MAP) framework. Finally, the optimal high-resolution face image is obtained by a global linear smoothing operator. It is can be seen from the experimental result that oriented facial features in the high-resolution face are recovered well. The most crucial is that our algorithm significantly reduces the computational complexity.
    No preview · Article · Jan 2010
  • Yanbing Xue · Wenhui Deng · Dongsheng Zhou · Hua Zhang
    [Show abstract] [Hide abstract]
    ABSTRACT: A single image super resolution algorithm for license plate preprocessing is proposed in this paper. The image to be enhanced is modeled as a Markov Random Field and is estimated from the input low resolution image by image patch pairs. From the input image and the training set, observation function and compatibility function can be calculated. Then Bayesian Belief Propagation is used to select the most probable high resolution patches candidate in the MRF model. The experiment shows that using this method can get better license plate with more information for further recognition.
    No preview · Conference Paper · Jan 2010
  • Yanjie Ma · Hua Zhang · Yanbing Xue
    [Show abstract] [Hide abstract]
    ABSTRACT: We address a novel method for super resolution based on Markov random field (MRF). Modeling image patches as MRF node, and we learn the parameters from training samples. Training sample set provide a candidate high-resolution interpretation for the low-resolution images. Given a new low-resolution image to enhance, we select from the training data a set of 10 candidate high-resolution patches for each patch of low-resolution image. In Bayesian belief propagation, we use compatibility relationships between neighboring candidate patches to select the most probable high-resolution candidate. The experimental results show that this method can obtain the better result.
    No preview · Article · Oct 2009
  • YanJie Ma · Hua Zhang · Yanbing Xue · Simiao Zhang
    [Show abstract] [Hide abstract]
    ABSTRACT: We address a learning-based method for super resolution. Training sample set provide a candidate high resolution interpretation for the low-resolution images. Modeling image patches as Markov network node, and we learn the parameters of the network from training set,compute probability distribution by K-means algorithm. Given a new low-resolution image to enhance, we select from the training data a set of 10 candidate high-resolution patches for each patch of low-resolution image. In Bayesian belief propagation, we use compatibility relationship between neighboring candidate patches to select the most probable high-resolution candidate. The experimental results show that this method can obtain better result.
    No preview · Conference Paper · Jan 2009
  • Degan Zhang · Hongyun Ning · Fayu Wang · Yanbing Xue
    [Show abstract] [Hide abstract]
    ABSTRACT: Apparently, this function of nomadic service is suitable for mobile services. But when nomadic service for computing task is realized among PC, laptop, or PDA, there are several difficult problems to be solved, such as how to supply the continuity and adaptability. In order to realize nomadic service, we design and improve relative approach. In this paper, firstly, formal description of task and its type have been given, then, classification of Agent and transferring granularity of task have been suggested, and then efficient mechanism of nomadic service based on agent for pervasive computing has been presented and designed, the description of suggested management platform supporting nomadic service
    No preview · Article · Oct 2008
  • Yanbing Xue · Hua Zhang · Fayu Wang · Xianbin Wen
    [Show abstract] [Hide abstract]
    ABSTRACT: A new algorithm in Exemplar-Based image completion is proposed using color Ratio Gradient. Color Ratio Gradient Histogram is robust to object occlusion and clustering which induce mismatch in general image match algorithm. The proposed algorithm can improve the quality of comparing similarity between source image patch and target image patch, then improve the quality of image completion. Experimentation results show that the proposed method improve the quality of image completion compared with previous algorithms.
    No preview · Conference Paper · Jun 2008