International Journal of Information Acquisition

This paper presents a method for detecting abnormal motion in real time using a computer vision system. The method is based on the modeling of human body image, which takes into account both orientation and velocity of prominent body parts. A comparative study is made of this method with other existing algorithms based on optical flow and the use of accelerometer body sensors. From the real time experiments conducted in the present work, the developed method is found to be efficient in characterizing human motion and classifying it into basic types such as falling, sitting, and walking. The method uses a Radial Basis Function Network (RBFN) to compute the severity coefficient associated with the type of motion, based on experience. The paper evaluates the various methods and incorporates the advantages of other methods in order to develop a more reliable system for abnormal motion detection.

The absence of GPS underwater makes navigation for autonomous underwater vehicles (AUVs) a challenge. Moreover, the use of static beacons in the form of a long baseline (LBL) array limits the operation area to a few square kilometers and requires substantial deployment effort before operations. In this paper, an algorithm for cooperative localization of AUVs is proposed. We describe a form of cooperative Simultaneous Localization and Mapping (SLAM). Each of the robots in the group is equipped with an Inertial Measurement Unit (IMU) and some of them have a range-only sonar sensor that can determine the relative distance to the others. Two estimators, in the form of a Kalman filter, process the available position information from all the members of the team and produce a pose estimate for every one of them. Simulation results are presented for a typical localization example of three AUVs formation in a large environment and indirect trajectory. The results show that our proposed method offers good localization accuracy, although a small number of low-cost sensors are needed for each vehicle, which validates that it is an economical and practical localization approach.

This paper introduces an efficient Eigen values based technique for online iris image compression and identification of a human including the case of identical twins. The iris image is extracted after removing the pupil, eye brow, skin and other noise disturbances from an actual image. The extracted iris image is divided into different blocks of size of 16 × 16. Now, Eigen values are calculated for each block and these Eigen values are stored in the smart card memory for further identification.

Therefore, when checking if two iris images are identical or not, all we need is to compare the stored Eigen values with online calculated Eigen values. If two iris images have the same Eigen values, this means that both iris images belong to the same person. In our research, we have concluded that iris images of different persons have different Eigen values, including the case of identical twins. We conducted experiments on CASIA and Multimedia University iris image databases and we found that our Eigen Values Based Iris Image Identification Technique is giving 99.99% accuracy for the same image of identical twins and individuals. The implementation leads us to believe that our method is giving the best matching result in the case of identical twins and individuals. It is an efficient, secure and economically feasible approach for online personal identification.

A capacitive MEMS Ultra-Low-Power readout for accelerometers and strain sensors using VerilogA models is presented. The VerilogA model of the accelerometers and strain sensors allows the simulation of a system in a half-bridge configuration. The gain of the system is controlled by integrating pulses from the excitation voltage which accurately controls the Signal-to-Noise ratio. A Figure-of-Merit of was achieved for a sensor range of ±2.0 g and ±20,000 με over a 100 Hz bandwidth. Residual motion artefacts are also canceled by the system.

Data mining methods are used to mine voluminous data to find useful information from data. The data that is to be mined may have a large number of dimensions, so the mining process will take a lot of time. In general, the computation time is an exponential function of the number of dimensions. It is in this context that we use dimensionality reduction techniques to speed up the decision-making process. Dimensionality reduction techniques can be categorized as Feature Selection and Feature Extraction Techniques. In this paper we compare the two categories of dimensionality reduction techniques. Feature selection has been implemented using the Information Gain and Goodman–Kruskal measure. Principal Component Analysis has been used for Feature Extraction. In order to compare the accuracy of the methods, we have also implemented a classifier using back-propagation neural network. In general, it is found that feature extraction methods are more accurate than feature selection methods in the framework of credit risk analysis.

This paper discusses the deployment of Genetic Algorithm optimization method for the synthesis of antenna array radiation pattern in adaptive beamforming. The synthesis problem discussed is to find the weights of the Uniform Linear Antenna array elements that are optimum to provide the radiation pattern with maximum reduction in the sidelobe level. This technique proved its effectiveness in improving the performance of the antenna array.