Search

Chaste and stationary foreground may occur in traditional background subtraction when objects start or stop moving. Eliminating ghosts and extracting stationary foreground immediately are crucial for improving the subsequent tasks such as object tracking, recognition and activity analysis. In this paper, we propose a method to detect ghosts and stationary foreground by dual-direction background modeling. The forward background model and the backward background model are built by GMM and a simple regression model respectively, which can detect not only the moving foreground but also the stationary foreground and the ghosts. Extensive experiment results demonstrate that the proposed algorithm is effective and efficient in eliminating ghosts and detecting stationary foreground.

In this paper, a recursive multiple objects tracking method based on data association is proposed. Firstly, an adaptive tracking window is designed for the scale changes of the object in motion scene. Then a data association method is proposed for multi-object tracking. The tracking window and the observation data of the same object at different time as well as the different objects at the same time are all associated to achieve robust tracking of the objects with different number and size. Experimental results show that the proposed multi-object tracking mechanism can deal with complicated situations such as object occlusion, merging, separating, appearing and disappearing.

Particle Filter is one of the most widely used algorithm in object tracking, because it can handle the nonlinear and/or non-Gaussian problems. However it fails when the tracked object is occluded by other objects. In order to solve such problems, in this paper, a memory-based state estimation scheme is introduced into the particle filter to estimate the state of the occluded object. When occlusion occurs, the tracked object is located by the states stored in the memory space. And the tracked object can be correctly relocated when recovered from the occlusion. The experimental results show the effects of our proposed method.

In this paper, a memory-based Gaussian Mixture Model (MGMM) is proposed inspired by the way human perceives the environment. The human memory mechanism is introduced to model the background, which can make the model remember what the scene has ever been and help the model adapt to the variation of the scene more quickly. Experimental results show the effect of the memory mechanism in segmenting moving objects with sudden partial changes in the background scene.

Human face detection is a key technology in face information processing, the speed of which is very important during real-time face detection for input images or input video sequences. This paper presents a novel face window searching algorithm based on evolutionary agent when detecting faces in gray-scale images. It can quickly And the candidate face windows through the evolutionary computation of distributed agents, which represent different kind of windows. Experimental results show that the evolutionary agent-based searching algorithm can increase the detection speed by 5–7 times compared with the traditional exhaustive searching method used in some general algorithms.

When tracking the moving targets in video image sequences with the existing particle filter, usually the tracking performance is not satisfactory due to the particle degradation and particle diversity loss. In this paper, we propose a novel particle filtering algorithm. In the algorithm, the multi-agent co-evolutionary mechanism is introduced into the particle re-sampling process and make the particle become an agent having ability of local perception, competitive selection and self-learning by the redefinition of particle agent and its local living environment. The re-sampling process is accomplished by the co-evolutionary behaviors among particles such as competition, crossover, mutation and self-learning, etc. It can not only ensure the particle validity but also increase the particle diversity. Experimental results show that the proposed algorithm can achieve better performance when tracking objects in complex video scenes.

Male meiosis is a specialized type of cell division that gives rise to sperm. Errors in this process can result in the generation of aneuploid gametes, which are associated with birth defects and infertility in humans. Until now, there has been a lack of a large-scale identification of proteins involved in male meiosis in mammals. In this study, we report the high-confidence identification of 3625 proteins in mouse male germ cells with 4C DNA content undergoing meiosis I. Of these, 397 were found to be testis specific. Bioinformatics analysis of the proteome led to the identification of 28 proteins known to be essential for male meiosis in mice. We also found 172 proteins that had yeast orthologs known to be essential for meiosis. Chromosome distribution analysis of the proteome showed underrepresentation of the identified proteins on the X chromosome, which may be due to meiotic sex chromosome inactivation. Characterization of the proteome of 4C germ cells from mouse testis provides an inventory of proteins, which is useful for understanding meiosis and the mechanisms of male infertility.

Gaussian Mixture Model (GMM) is one of the best models for modeling a background scene with gradual changes and repetitive motions. However, it fails in segmenting moving objects when the scene changes sharply. To handle this problem, an improved background modeling algorithm-Intelligent GMM (IGMM), which is inspired by the way human perceive the environment to tackle sharp changes in the scene, is proposed. In addition, each foreground pixel is relabeled according to its neighbourhood information in the binary foreground image to effectively reduce the number of False Negatives (FNs). The proposed method can make the GMM remember or forget what the scene has ever been during the learning and updating period. Experimental results show that it can help segmenting moving objects precisely when the scene changes sharply and improving the robustness of the scheme.

Network violence can be divided into network symbol violence, the network trial, and group violence, the cause of which includes the hidden characteristics of networks, collective behavior and entertainment desire, the dislocation of freedom, and the deficiency of law and moral restriction. The prevention and cure strategies of network violence include intensifying network management, implementing network real-name, strengthening the construction of network legal system and educating the youth.