Coupled neural P (CNP) systems are a recently developed Turing-universal, distributed and parallel computing model, combining the spiking coupled mechanisms of neurons. This paper focuses on how to apply CNP handle fusion multi-modality medical images proposes novel image method. Based two with local topology, an framework in nonsubsampled shearlet transform (NSST) domain is designed, where used control low-frequency NSST coefficients. The proposed method evaluated 20 pairs compared seven...

In this article, a cascade fuzzy neural network (FNN) control approach is proposed for position of quadrotor unmanned aerial vehicle (UAV) system with high coupling and underactuated. For the attitude loop limited range, FNN controller parameters were trained offline using flight data, whereas loop, method based on compensation proportional-integral-derivative (PID) was adopted to tune online adaptively. This not only combined advantages systems networks but also reduced amount calculation...

This paper focuses on a parallel version of particle swarm optimization (PSO) algorithm which can significantly reduces execution time for solving complex large-scale problems. gives an overview PSO algorithm, and then proposes design implementation PSO. The proposed eliminates redundant synchronizations optimizes message transfer to overlap communication with computation. experimental results showed that 13.2 times speedup was obtained by the 14 processors.

Based on the sharing features of a variety license plates (LP), vertical edge was first detected by Sobel detector. Then, some approaches were adopted to remove invalid regarding characteristics grayscale jump and density, so that regions having LP preserved. Next, horizontal projections mathematical morphology (MM) operation, region searched. color-reversing judgement conducted color analysis, binarization done based core in LP. Afterward, characters segmented means prior knowledge...

In this paper, an effective approach to vehicle license plate recognition based on Extremal Regions (ERs) and Self-adaptive Evolutionary Extreme Learning Machine (SaE-ELM) is proposed. the detection step, some computations including morphological operations, various filters, different contours validations are sequentially performed extract image regions as candidate plates. Then, accurate character segmentation achieved through a proper selection of ERs. HOG (histogram oriented gradients)...

In today’s society, graphic design, as a popular image processing technology, plays an increasingly important role in people’s lives. the specific operation process of It is no longer restricted to traditional development mode, such file format and other factors. With computer network people promote design by constructing color management system. At same time, construction system can help change colors define when they information output pictures. printing pictures, order make used clearly...

Fast 3D reconstruction of tool wear from 2D images has great importance to measuring and objective evaluating condition, determining accurate change insuring machined part's quality. Extracting information zone based on monocular multi-color structured light can realize fast recovery surface topography wear, which overcomes the problems traditional methods such as solution diversity slow convergence when using SFS method stereo match multiple images. In this paper, a kind new illuminator was...

A novel image recognition method based on chaotic-particle swarm-optimization-neural network algorithm was presented. The chaotic mapping mechanism and particle swarm were used to optimize the weight threshold of BP neural which applied image. simulation results show this new can overcome problems that is easy fall into local optimum sensitive initial value, has better rate stronger robustness.

Visible and near-infrared(NIR) band sensors provide images that capture complementary spectral radiations from a scene. And the fusion of visible NIR image aims at utilizing their spectrum properties to enhance quality. However, currently algorithms cannot well take advantage properties, as lack information complementarity, which results in color distortion artifacts. Therefore, this paper designs model level physical signals. First, order distinguish between noise useful information, we use...

With the development of distributed power generation, fluctuation its output may bring a higher risk serious cascading failure in an urban grid with change natural conditions. The existing assessment methods fail to consider generation. Therefore, this paper proposes method for failures grids including generation and electric vehicles. Firstly, based on probabilistic models wind power, photovoltaic, vehicles, random flow algorithm semi-invariants is applied solve flow. And then, fault chains...

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Fully online learning has become a common option in many universities worldwide the post-COVID-19 era. The study aimed to evaluate dimensions and characteristics of fully self-efficacy among Chinese undergraduate medical nonmedical students. A cross-sectional was conducted from January August 2023 at Xihua University China. stratified cluster sampling method used enroll participants version scale (OLSES) collect demographic information. Cronbach's alpha coefficient, exploratory factor...

Due to the limitation of image acquisition equipment itself and constraints taking environment other factors, resolution contrast obtained images have declined, some important information in was hidden process. Therefore, order get a richer picture, source must be enhanced prior integration, so as access information. The proposed paper introduces general integration scheme based on enhancement, presents fusion method with wavelet enhancement technology. We first enhance according technology,...