Person re-identification has long been a significant research direction in intelligent network surveillance. The challenging issues in person re-identification consist in pose, viewpoint and illumination changes and occlusions. In this paper, we propose a saliency based approach, which simulates the recognition process of the human brain, to tackle these issues. When people see a picture, they tend to focus on the salient areas and information in those areas is more determinant in the further matching and identification process. This so-called visual attention mechanism has long been studied and used in image segmentation, tracking, detection and recognition. To simulate this distinctive mechanism, we first calculate the saliency map which indicates the conspicuity of each pixel, and then we extract the saliency map weighted HSV histograms by giving each pixel a weight according to its saliency. We also design another feature, the salient colors, to address the occlusion problem. By opportunely combining these two features, our approach achieved state of the art performances.
In this paper, we focus on the problem of detection and localization of crowd escape anomalous behaviors in video surveillance systems. The scheme proposed can not only detect the abnormal events which have been studied, but also detect the possible location of abnormal events. People usually instinctively escape from a place where abnormal or dangerous events occur. Based on this inference, a novel algorithm of detecting the divergent center is proposed: The divergent center indicates possible place where abnormal events occur. The model of crowd motion in both the normal and abnormal situations has been made according to the proposed method. Intersections of vector are obtained through solving the straight line equation sets, where the straight line Equation sets are determined by the location and direction of motion vector which are calculated by the optical flow. Then the dense regions of intersection sets, i.e., the divergent center, are obtained by using the distance segmentation method, the threshold method and the graphical method. Escape detection is finally judged according to the speed and energy of motion and the divergent center. Experiments on UMN datasets and other real videos show that the proposed method is valid on crowd escape behavior detection.
There are quantity differences among multi-source domains and data distribution differences between source domain and target domain in multi-source domain adaptation, which makes it difficult to extract transferability features among all domains and makes the classification performance of the target domain poor. To address this problem, this paper proposes an image classification method based on sequential multi-source domain adaption method (SMSDA). A source domain arrangement mechanism is proposed, which takes the distribution differences and quantity differences among domains as description metrics. The target domain can perform domain adaptation with the source domain sequentially, which solves the problem that the classification performance varies greatly under different domain adaption orders of the source domain and target domain. Meanwhile, SMSDA proposes a multi domains feature local alignment rule through the introduction of fine-grained idea. It perfects the low transfer performance in the global distribution alignment. Furthermore, an adaptive adjustment strategy of sample weight is designed to improve the low classification accuracy under class-imbalance. Finally, the experimental results on three benchmark datasets show that the proposed method performs better in classification accuracy compared with the conventional methods.
We in this paper give a decomposition concerning the general matrix triplet over an arbitrary divisionring F with the same row or column numbers. We also design a practical algorithm for the decomposition of thematrix triplet. As applications, we present necessary and suficient conditions for the existence of the generalsolutions to the system of matrix equations DXA = C1, EXB = C2, F XC = C3 and the matrix equation AXD + BY E + CZF = Gover F. We give the expressions of the general solutions to the system and the matrix equation when thesolvability conditions are satisfied. Moreover, we present numerical examples to illustrate the results of thispaper. We also mention the other applications of the equivalence canonical form, for instance, for the compressionof color images.
The next generation network is a type of applicationoriented one,and service creation and deployment are of overwhelming importance. Softswitch architectures, such as Parlay, utilise service logic decomposition to achieve it. These architectures decompose service logic into servicedependent and serviceindependent parts. Service logic decomposition is applied to the functionally rich TINA service components resulting in generic and reusable software subcomponents,and reuse service logic during service creation thus ensuring rapid deployment of third party services in the NGN.
In this paper, we propose a novel algorithm based on the acceleration feature to detect anomalous crowd behaviors in video surveillance systems. Different from the previous work that uses independent local feature, the algorithm explores the global moving relation between the current behavior state and the previous behavior state. Due to the unstable optical flow resulting in the unstable speed, a new global acceleration feature is proposed, based on the gray-scale invariance of three adjacent frames. It can ensure the pixels matching and reflect the change of speed accurately. Furthermore, a detection algorithm is designed by acceleration computation with a foreground extraction step. The proposed algorithm is independent of the human detection and segmentation, so it is robust. For anomaly detection, this paper formulates the abnormal event detection as a two-classified problem, which is more robust than the statistic model-based methods, and this two-classified detection algorithm, which is based on the threshold analysis, detects anomalous crowd behaviors in the current frame. Finally, apply the method to detect abnormal behaviors on several benchmark data sets, and show promising results.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)