Papers by Author: Hong Yu Wang

Paper TitlePage

Authors: Ming Xin Jiang, Hong Yu Wang, Chao Lin
Abstract: As a basic aspect of computer vision, reliable tracking of multiple objects is still an open and challenging issue for both theory studies and real applications. A novel multi-object tracking algorithm based on multiple cameras is proposed in this paper. We obtain the foreground likelihood maps in each view by modeling the background using the codebook algorithm. The view-to-view homographies are computed using several landmarks on the chosen plane. Then, we achieve the location information of multi-target at chest layer and realize the tracking task. The proposed algorithm does not require detecting the vanishing points of cameras, which reduces the complexity and improves the accuracy of the algorithm. The experimental results show that our method is robust to the occlusion and could satisfy the real-time tracking requirement.
Authors: Xiu Ying Zhao, Hong Yu Wang, Cheng Zhi Yang, Hong Chao Wu
Abstract: This paper presents a new algorithm for the deinterleaving of radar signals, using the direction of arrival (DOA), carrier frequency (RF), and time of arrival (TOA). The algorithm is mainly applied to pulse repetition interval (PRI) signals. This algorithm consists of two steps: In the first step, a PRI transformation is used to the received pulses after pre-deinterleaved of frequency and DOA. In this step, radar signals having the same frequency and DOA are identified as the same class. In the second step, the number of existing emitters and their PRIs is determined by using TOA information. The algorithm for deinterleaving uses the information obtained from the previous analysis to reduce the PRI errors. The simulation results show that the algorithm is successful in high pulse density environments and for the complex signal types.
Authors: Wen Zhu Sun, Hong Yu Wang, Jie Wang
Abstract: This paper presents a layered error resilient scheme for Set Partitioning in Hierarchical Trees (SPIHT) coding which incorporates the Multiple Description Coding (MDC) into the data partition. The major feature of this scheme is that the MDC is adopted in the process of packetizing the spatial orientation trees, so as to protect the important information bits according to different importance and channel state. The proposed scheme can allocate redundancy accurately and flexibly, by differentiating importance of bits through different bit planes. Experimental results show that the proposed method achieves good performance over packet erasure channels.
Authors: Ming Xin Jiang, Hong Yu Wang
Abstract: Reliable tracking of multiple people in cluttered or complex situations is a challenging visual surveillance problem since the high density of objects results in occlusion. In order to deal with this problem, multiple synchronized cameras were mounted at various heights in our experiment. To ensure the existence of the homography, it is necessary to assume that different views share a common dominant ground plane. Thus, corresponding people have to be located within the multi-camera surveillance system, accurate multi-people localizing is an important prerequisite to reliable tracking. In this paper, we present a novel approach to fusing foreground information on planes of different height from multiple views to increase accurateness of localization. Our method does not require fully camera calibration. First we obtain the foreground likelihood maps in each view by modeling the background using codebook algorithm. Then we compute homographies induced by multiple planes and obtain the localization of multiple people at multiple planes. Finally, we adopt an algorithm of feature-based using Kalman filter motion to handle multiple objects tracking at multiple plane. The experimental results show that our method is valid and has nice robustness to the occlusion in crowed environments.
Authors: Ming Xin Jiang, Xing Yang Cai, Hong Yu Wang
Abstract: An early smoke detection algorithm based on Codebook model and multiple features is presented in this paper. First, the foreground is obtained by using the Codebook algorithm. Second, the model of color distribution and the model of shape feathers of smoke are applied to detect the suspected smoke area in the foreground. Finally, the false alarm rate is reduced effectively by using dynamic features in the diffusion process of smoke. Experimental results show that our algorithm has good detection performance and achieves real-time requirement which is very important for real application.
Authors: Xiu Ying Zhao, Hong Yu Wang, Shou Yu Tong, De You Fu, Hai Shen Zhou
Abstract: The spectral subtraction is one of the best methods for elimination of approximate cyclical engine’s noise from degraded speech signal. Here we turn to research about the nonlinear spectral subtraction method and its improved model. After studying the nonlinear method we turn to this method that whether it can improve the quality of enhanced speech signal, propose the short-time spectral subtraction, which needs two inputs. The main input is containing the voice that is corrupted by noise. The other input (noise reference input) contains noise related in some way to that of the main input (background noise). Then use the main input’s frequency spectrum subtract the other input’s frequency spectrum. The results of experiment have proved it’s effective.
Authors: Kai Yan Zhu, Hong Yu Wang, Da Xing Qian
Abstract: In this paper, we consider a space-time and amplify-and-forward(ST-AF) cooperative system which consists of two-antenna source, single-antenna relay and destination. Source transmits Alamouti space-time coding symbols to destination with cooperation of relay which adopts AF strategy. Closed-form symbol error rate(SER) is derived for the ST-AF system with PSK signals. Moreover, an SER approximation are developed to show the asymptotic performance of the ST-AF cooperative system in medium and high SNR regimes. For comparison, the SER approximation of another cooperative space-time coding(C-STC) scheme is also derived. Theoretical analysis shows that the ST-AF can obtain more diversity gain and achieve higher diversity order than C-STC. Finally, computer simulations validate the theoretical analysis.
Authors: Zhi Jun Gao, Hong Yu Wang, Tie Chi, Bin Ma
Abstract: In this paper, we researched the monitoring technology of structural health which based on wireless sensor networks, and combined with the characteristics of structures, we designed a wireless sensor network monitoring systems, then the composition and topology of wireless sensor networks was proposed. At last, we used the improved ant colony algorithm to develop network energy-saving strategy which parameters are the energy of wireless sensor nodes, resized using node energy to control the routing between nodes, effectively reducing the energy consumption of wireless sensor monitoring network. The simulation result proved that the algorithm is effective, reliable and energy-saving effect.
Authors: Da Xing Qian, Hong Yu Wang, Fang Lin Niu
Abstract: The bit stream extraction plays an important role in Scalable Video Coding (SVC) [1]. However, one downside of current video coding methods is to ignore the video contents which is in fact an important factor for video coding efficiency. Therefore, an equivalent MSE method is proposed in this paper to extract substreams in the temporal and spatial enhancement layers. When the Motion Vectors (MVs) are large in one video, a larger frame rate is necessary to maintain the continuity of the object movement which makes no jump in the visual sense. In this sense, substreams extraction in temporal enhancement layer has to be satisfied. On the other hand, if there are some larger high-frequency components in a single frame of the video, that is to say, there are some higher spatial details in the video stream. As a result, it should try to meet the extraction requirement in spatial enhancement layer. This method has the advantage of considering the contents of the video, which can effectively improve the coding performance and quality. The experimental results have demonstrated the improved quality of reconstructed video for the equivalent MSE method when extracting bit stream arbitrarily at the same bandwidth.
Authors: Xiu Ying Zhao, Hong Yu Wang, De You Fu, Hai Shen Zhou
Abstract: The presence of noise superimposed on a signal limits the receiver’s ability to correctly identify the intended signal. The principal of adaptive noise cancellation is to acquire an estimation of the unwanted interfering signal and subtract it from the corrupted signal. Noise cancellation operation is controlled adaptively with the target of achieving improved signal to noise ratio. This paper describes the Least Mean Squares (LMS) adaptive filtering algorithm. The algorithm was implemented in Matlab and was tested for noise cancellation in speech signals.
Showing 1 to 10 of 10 Paper Titles