Applied Mechanics and Materials Vols. 687-691

Abstract: With the development of computer and embedded technology, ubiquitous computing age is coming. In the ubiquitous age, people could get all kinds of services anywhere and at any time. In this paper, we design an indoor location system based on ultrasonic sensor and radio frequency technology. The ultrasonic positioning technology has many merits of high positioning accuracy, simple circuit structure and low cost. There are some components such as locating system, communication system, location system, security system and so on, emphasized supporting technologies in order to provide researchers with general cognition. The principle and the design of the system are described in detail. The system consists of the single chip circuit, the circuit of ultrasonic, the circuit of radio frequency signal, communication interface circuit and so on. Perceptual system can promote environmental intelligence and human services development, but also can promote public places security monitoring and early warning intelligence and automation.

Abstract: Full-service operations competition is going on among Chinese telecom operators. How to offer integrated information service and acquire competitive advantage has become a new research topic. So this paper focuses on the integrated information service, proposes the eco-industrial value chain oriented view for telecom operator and builds the model of integrated information service quality based on SERVQUAL to help China telecom operators make the integrated information service strategies and tactics in dynamic eco-industrial value chain.

Abstract: According to the trend of the intelligent monitoring system, on the basis of the study of gait recognition algorithm, the intelligent monitoring system is designed based on FPGA and DSP; On the one hand, FPGA’s flexibility and fast parallel processing algorithms when designing can be both used to avoid that circuit can not be modified after designed; On the other hand, the advantage of processing the digital signal of DSP is fully taken. In the feature extraction and recognition, Zernike moment is selected, at the same time the system uses the nearest neighbor classification method which is more mature and has good real-time performance. Experiments show that the system has high recognition rate.

Abstract: Detecting tampered image is a challenging work due to the high volume of image database and the accurate definition of tampering. We propose a novel algorithm based on standard deviation which could detect the tampered automatically, furthermore, localization and extraction process is conducted to optimize the proposed method. Color reduction technique, intensity based method for edge detection and horizontal based localization approach are applied here to fulfill the algorithm. The core idea of the paper is that normally tampered regions process high standard deviation while compared with non-tampered areas. As the result, the output of our algorithm is tampered regions. By presenting promising experience, the performance of proposed method is analyzed. Further application and possible optimization are discussed.

Abstract: Road detection is a primary problem for autonomous vehicle. There have been many approaches attempting to solve this problem. However, most of the approaches tend to be affected by shadow, occlusion and not always robust. In this paper, we propose a new approach to detect the road boundary, our approach has two parts: firstly, we detect the road boundary by random forests on the pixel level, then we track the road boundary by particle filter with the parameters of vanish point position and road boundary slope. The experiment shows our approach is an effective way to detect road boundary fast and robustly.

Abstract: This paper proposed a target detection method in the presence of multiple jamming, which may come from mainlobe or sidelobe of the array antenna. The jamming direction-of-arrival (DOA) is estimated using MUSIC algorithm, and a multiple jamming cancellation beamformer is suggested, which is based on a FIR filer in space domain, the beamformer is followed by a pulse compression filter in frequency domain which can improve the resolution in time domain. We take a simulation experiment for our method, considering a digital array radar with pulse compression technique, the experiment indicate that the target detection method have a good performance and can be used in digital array radar system.

Abstract: Apply driving assistance system to vehicles can significantly reduce accidences and thus attracts much interest nowadays. However, most existing systems are designed specifically for small vehicles and always suffer from drawbacks such as low pedestrian and vehicle detection accuracy and long detection time. To solve these issues, in this paper we develop a driver assistance system based on radar and camera, which can be applied to large vehicles and can detect vehicle and pedestrian simultaneously. Specifically, we combine the image subtraction technique and histogram algorithm to perform pedestrian and vehicle detection to improve detection rate. What’s more, this system can automatically determine whether the object is inside a danger region. If yes, an associated warning signal will be triggered to alarm the driver. Experimental results show that the successful detection rate is sufficiently good and the detecting speed is fast enough to timely alarm the driver to avoid accidents.

Abstract: With the raise the level of sports competition, sports training requirements is also rising, just by virtue of the coaches of the intuition training way already can't satisfy the stages of athletic training needs, in order to adapt to this situation, the computer vision technology is increasingly cited used in sports training, because the accuracy and memory of the machine vision has better compared with the human eye, can capture the target quickly and efficiently, and can be a variety of motion data record goal, to provide more theoretical, data description for the athlete's movement. The level of theory mainly includes three parts: basic feature layer, model, event layer and knowledge layer. On the theoretical level, corresponding to the corresponding technical route: basic motion feature extraction, pattern, event mining technology. The theoretical level in each layer are not independent, low layer is tall service, but they also have their own characteristics, can be used for mining information.

Abstract: In order to improve the speed of collision detection between objects in the large-scale and complex scene, this paper proposed an improved collision detection algorithm based on GPU, In this method, we first divided the virtual space into several grids to rule out the impossible intersecting objects rapidly using the GPU acceleration technology; secondly, we adopted parallel technology to build K - DOP bounding boxes for the objects in the same grids and then detected whether the K - DOP bounding boxes intervene or collide to conform the potential colliding primitive pairs; Finally we traveled the final triangle intersection tests on GPU. Compared to the traditional K-DOP compared bounding box collision detection, The algorithm can effectively improve the real-time collision detection.

Abstract: The selection of kernel function in Support Vector Machine (SVM) has a great influence on the model performance. In the paper, Mexico hat wavelet kernel is introduced to employ the kernel function of SVM, and theoretically it has be prove that, Mexico hat wavelet kernel satisfies the Merce condition, that is the necessary condition as the kernel function of SVM. Simulation on the anomaly detection shows that the capability of SVM based on Mexico hat wavelet kernel is better than that of SVM based on RBF kernel with a satisfactory result for anomaly intrusion detection.