Papers by Keyword: Information Fusion

Paper TitlePage

Authors: Guo Dong Zhang, Rui Min Qi
Abstract: In the field of information confusion, evidence theory takes advantage of its uncertainties. But in practical evidences are always mutually independent, However, multi-source information fusion is a comprehensive integration and then obtaining decision-making, sometimes the result of information fusion give us wrong conclusion. So an improved information fusion algorithm is proposed in this paper. It can heighten the information confusion reliability and accuracy in a practical example.
1909
Authors: Hong Sheng Tan, Xia Ming Yuan, Ji Hong Zhu
Abstract: For the angular velocity gyro drift and noise and the inertial navigation solution of the integral nature, the navigation error of the attitude will accumulate over time. With the increase of navigation time, the system may fail. In high dynamic motion the accelerometer / magnetometer attitude measurement navigation system will not work properly. Therefore, the accelerometer / magnetometer attitude determination system, haven’t no drift, but poor dynamic performance. This article will realize a low-cost, precision Attitude Calculation System with more effective integration of the two.
802
Authors: Abasi
Abstract: With the rapid development of computer technology and communication technology, user demand more and more and with the application of computer network becoming more and more widely, its scale is becoming increasingly big. But on the other hand, there are so many network security events and the traditional single defense equipment or testing equipment has been unable to meet the demand of security; computer network is faced with severe situation of the information security. Network security situation assessment technology can reflect on the whole dynamic network security situation, and give early warning to its development trend by comprehensively considering various security factors. Therefore, the network security situation assessment model and key technology has become the hot topic of research in the field of network security.
5312
Authors: Jin Liu, Cang Ming Liu, Ting Ting Liu
Abstract: Starting from the goal to overcome the limitation of the single type algorithm of iris recognition. In this paper, a two-step iris recognition algorithm is proposed based on feature extraction on corner and Gabor transform .First of all, it uses the corner feature of iris texture to recognize for achieving the purpose of removal of alien and confirm the vast majority of similar samples rapidly; if it unabes to determine the category in the first step, the result of the first step is used as the initial angle in normalized phase, and then it carries out iris feature extraction and recognition based on Gabor transform. The experimental results show that this results the speed and performance in the recognition improved significantly than the traditional algorithm of single feature recognition.
559
Authors: Li Fu, Jun Xiang Wang
Abstract: A design and implementation of a detection system for dangerous driving was proposed based on multi-sensor-fusion. It is actually an embedded system consisting of visual,sensor, acceleration sensor, alcohol sensor input, and ARM cortex-M3 microcontroller. Experiment results show that the system has high linearity, high sensitivity,and excellent real-time performance. It can be further used to validate the multi-sensor information fusion algorithms in the field for improving the low reliability of the current detection by using one single-sensor method
91
Authors: Shi Lin Zhang, He Ping Li, Shu Wu Zhang
Abstract: In this paper, we present a method for video semantic mining. Speech signal, video caption text and video frame images are all key factors for a person to understand the video content. Through above observation, we bring forward a method which integrating continuous speech recognition, video caption text recognition and object recognition. The video is firstly segmented to a serial of shots by shot detection. Then the caption text and speech recognition results are treated as two paragraphs of text. The object recognition results are presented by bag of words. The above three aspects of texts are processed by part of speech and stemming. Then only the noun words are kept. At last a video is represented by three bags of words. The words are further depicted as a graph. The graph vertices stand for the words and the edges denote the semantic distance between two neighboring words. In the last step, we apply the dense sub graph finding method to mine the video semantic meaning. Experiments show that our video semantic mining method is efficient.
64
Authors: Yu Tao, Yong Chao Wei
Abstract: One framework that fuses features for abnormal behavior recognition in videos was proposed, which provides increased robustness to noise and pose variation. First, ROI division of suspected pedestrian is achieved by optical flow method, which greatly reduces the dimensions of training data. After that, the proposed method uses the adaboost training to recognize pedestrian effectively. To capture the imaging variations and attributes of individuals, we use types of features: center track and inclination angle. Finally, D-S evidence theory is used to combine these features, which aims to recognize the abnormal behavior correctly. The results demonstrate that the recognition rate can be improved by the fusion of features.
1949
Authors: De Min Li, Jian Zou, Kai Kai Yue, Hong Yun Guan, Jia Cun Wang
Abstract: Evacuation for a firefighter in complex fire scene is challenge problem. In this paper, we discuss a firefighters evacuation decision making model in ad hoc robot network on fire scene. Due to the dynamics on fire scene, we know that the sensed information in ad hoc robot network is also dynamically variance. So in this paper, we adapt dynamic decision method, Markov decision process, to model the firefighters decision making process for evacuation from fire scene. In firefighting decision making process, we know that the critical problems are how to define action space and evaluate the transition law in Markov decision process. In this paper, we discuss those problems according to the triangular sensors situation in ad hoc robot network and describe a decision making model for a firefighters evacuation the in the end.
504
Authors: Xiao Yin Xu, Li Hong Ren, Yong Sheng Ding
Abstract: In this paper, we use GA to improve the D-S evidence theory, and apply the improved D-S evidence theory to VIP intelligent recognition and recommendation system. In the VIP intelligent recognition and recommendation system of clothes, there are three main evidences: body size, personal preferences, and purchase records. So collision often happens inevitable. This requirement asks us to find out a suitable method to identify the VIPs needs. D-S evidence theory can improve the rate of identification, but has no idea about the collision. The improved D-S evidence theory based on genetic algorithm can deal with the collision evidence and improve the rate of the identification and the stability. As such we can provide VIP more suitable recommendation. The experiment results of clothes recommendation demonstrate the flexibility of the improved method.
2442
Authors: Ming Hui Deng, Qing Shuang Zeng, Lan Ying Zhang
Abstract: In this paper, a robust image watermarking method with S-Hough transformation is proposed which is robust against geometric distortion. This watermarking is detected by a linear frequency change. The chirp signals are used as watermarks and this type of signals is resistant to all stationary filtering methods and exhibits geometrical symmetry. In the two-dimensional Radon-Wigner transformation domain, the chirp signals used as watermarks change only its position in space/spatial-frequency distribution, after applying linear geometrical attack. But the two-dimensional Radon-Wigner transformation needs too much difficult computing. We propose a modified Hough transformation to provide improved energy concentration of the Dopplerlet transform. The proposed scheme can resolve the time-frequency localization in a better way than the standard Dopplerlet transformation. The watermark is embedded in the 1D improved Dopplerlet transformation domains. The watermark thus generated is invisible and performs well in test and is robust to geometrical attacks. Compared with other watermarking algorithms, this algorithm is more robust, especially against geometric distortion, while having excellent frequency properties.
2211
Showing 11 to 20 of 148 Paper Titles