Maximum Search Algorithm Research Articles

Simultaneous Inversion of Multiwindow Surface Wave Dispersion Data in 2D Layered Media

Seismic Surface Wave Methods (SWMs) have been one of the effective tools for retrieving near-surface shear-wave velocity (Vs). However, a critical problem associated with the SWMs is that using a horizontally layered model assumption in inversion may introduce blurred errors into 2D Vs imaging due to the presence of lateral variations. In addition, the increase of investigation depth necessary for the SWMs may cause an increase in errors since a wider acquisition array is essential to acquire longer wavelength data precisely. In this study, we analyzed the relationship between the output power, calculated by an amplitude normalized frequency-domain beamforming and the local phase velocity following the ray-theory approximation. Based on this relationship, we calculated the theoretical dispersion curves of multichannel surface wave data using a local maximum search algorithm. Then, we presented an algorithm for simultaneously inverting surface wave dispersion data extracted from multi-size spatial windows. We parametrized the inverted model in a 2D layered pattern. Finally, we used synthetic and field data sets to test the effectivity of the proposed method. Both results showed that the lateral resolution of the Vs model retrieved by the simultaneous inversion is improved compared to the horizontally layered model-based inversion.

Intellectual Homography Based Global Motion Estimation For Blocks In Video Coding

Block matching techniques for estimation of motionis very commonly utilized for current video coding standards as it is simple and has reasonable performance. There always lies a homography among the consecutive frames in the sequences of video that is taken using the PTZ cameras because of the restricted movements. This symmetrical relationship of the frame is very useful to decrease the spatial redundancy. The estimation of motion is playing a very significant part in science nowadays. Many algorithms are been proposed to observe movement of the objects present in the space. In this project, a homographybased local maximum search algorithm is proposed for the accurate guidance of the small bodies present in the space. This method will helps for the accurate movement detection. The proposed method is proven to be more for the sequences when compared with the traditional fast algorithms. The simulation is carried out using MATLAB software.

New scheme for finite-retardation limitations of linear retarders with fixed axes in polarization control

Finite retardation ranges of linear retarders with fixed axes limit their applications in polarization control. In this work, we present a simple and efficient constraint scheme for this finite-retardation limitation. Its theoretical basis is given geometrically and mathematically. The new polarization control algorithm combines the constraint scheme and a widely-used maximum-search algorithm. Both simulations and experiments confirm the effectiveness and practicality of the proposed scheme and control algorithm. In experiments, the control system uses four cascaded linear retarders with fast axes alternately oriented at 0° and 45°. Each retarder (fiber squeezer) has a finite retardation range of 4π. For a 15-min test, the mean (maximum) polarization error angle is 0.09 (0.28) rad while stabilizing a polarization fluctuation at an average speed of 26rad/s. Furthermore, no events of hitting retardation limits occur.

Investigation of peak wavelength detection of fiber Bragg grating with sparse spectral data

A simple equation for peak wavelength detection of fiber Bragg grating (FBG) with three spectral data is proposed, which has almost the same order of computing load as the conventional maximum search algorithm and uncertainty as the least-squares fit algorithm. The factors effecting performance of the new technique was investigated by simulation. For FBG with spectral bandwidth 0.2 nm, simulation results show that sample resolution lying between 0.07 and 0.16 nm have small measure standard deviation, which is less than 1 pm for SNR greater than 20.75 dB. Experiments were carried out with FBG and interrogator based on InGaAs linear array, and the results verified the effectiveness of the new technique.