Search

For training the thought concentration ability of hyperactivity sufferers, this paper proposed a kind of electroencephalogram (EEG) control variable extraction method and the motion control strategy. Firstly, the technologies of EEG acquisition from frontal lobe, and the wireless data transmission from the acquisition card to the Pad Phone were realised. Then, through the wavelet Mallat algorithm and FFT to extract the EEG control variable, the accurate controls of start, stop, and running velocity of the toy rail car were implemented. Thirdly, the mathematical model between the velocity and the EEG control variable was established, and then the start threshold, the swerve control and the speed adjust method were accurately designed using this model, so that the delay start and the swerve speed were solved. In addition, the wireless data transmissions among the Pad Phone, the EEG acquisition module and the railelectric potential controller were realised. Finally, it was proved that the proposed EEG control variables and the control strategy effectively fulfil the accurate speed control and the stable motion control.

In this paper, the purpose is to develop a Short Message Service (SMS) system that can aid the ALS patients to control a mobile phone to send a message by using of the steady state visually evoked potentials (SSVEP) brain computer interface (BCI) to a person who can help him/her. The EEG signals may be detected by electrodes and signal extracting chip. Then these signal can be transmitted to the mobile platforms by using of Bluetooth interface. Finally, the message and phone number can be selected in accordance with the components in frequency domain transferred from time domain. The experimental results had shown that the proposed system can easily send a message from a mobile phone to another.

The key step for gesture recognition is hand gesture segmentation. A new dynamic thresholding segmentation approach named Adaptive Threshold Segmentation Algorithm (ATSA) was developed. The segmentation effect evaluation was conducted based on the construction of a new color space model of skin model. The images of hand gesture were processed by using the proposed ATSA and the Fixed Threshold Segmentation (FTS) algorithm. Comparing with FTS, the ATSA is experimentally demonstrated that the segmented result has a less brightness impact, a lower redundancy rate, a lower rate of false alarm and missing, and a higher integrity rate.

A new fuzzy Hopfield-model net based on rough-set reasoning is proposed for the classification of multispectral images. The main purpose is to embed a rough-set learning scheme into the fuzzy Hopfield network to construct a classification system called a rough-fuzzy Hopfield net (RFHN). The classification system is a paradigm for the implementation of fuzzy logic and rough systems in neural network architecture. Instead of all the information in the image being fed into the neural network, the upper- and lower-bound gray levels, captured from a training vector in a multispectal image, are fed into a rough-fuzzy neuron in the RFHN. Therefore, only 2/Npixels are selected as the training samples if an N-dimensional multispectral image was used. In the simulation results, the proposed network not only reduces the consuming time but also reserves the classification performance. © 2004 Society of Photo-Optical Instrumentation Engineers.

The fuzzy possibilistic c-means (FPCM) is embedded into a 2-D Hopfield neural network termed the fuzzy possibilistic Hopfield network (FPHN) to generate an optimal solution for vector quantization (VQ) in the discrete cosine transform (DCT) and the Hadamard transform (HT) domains. The information transformed by DCT or HT is separated into dc and ac coefficients. Then, the ac coefficients are trained using the proposed methods to generate a better codebook based on VQ. The energy function of the FPHN is defined as the fuzzy membership grades and possibilistic typicality degrees between training samples and codevectors. A near global-minimum codebook in the frequency domains can be obtained when the energy function converges to a stable state. Instead of one state in a neuron for the conventional Hopfield nets, each neuron occupies two states called the membership state and the typicality state in the proposed FPHN. The simulated results show that a valid and promising codebook can be generated in the DCT or HT domains using the FPHN. © 2002 Society of Photo-Optical Instrumentation Engineers.

Junctions convey rich information and are often used to identify objects in the scene or for stereoscopic matching or displacement vector measuring. In a previous work, an accurate wavelet-based corner detection method has been proposed. This work is an extension of the work on junction characterization and detection in real images. We first perform an analytical study for a general junction model that allows us to better understand the behavior of junctions in the wavelet domain. Two new properties, including the scale proportion property of the modulus and the scale invariant property of orientation in the wavelet domain, are derived. These properties are applied to accurately detect junctions. Several promising experimental results have been carried out using noisy and noisy-free real images. These results show that our approach is a general methodology for accurate detection of discrete surface intersection. © 2002 Society of Photo-Optical Instrumentation Engineers.

An unsupervised parallel approach called Annealed Chaotic Competitive Learning Network (ACCLN) for the optimization problem is proposed in this paper. The goal is to modify an unsupervised scheme based on the competitive neural network using the chaotic technique governed by an annealing strategy so that on-line learning and parallel implementation to find near-global solution for image edge detection is feasible. In the ACCLN, the edge detection is conceptually considered as a clustering problem. Here, it is a kind of competitive learning network model imposed by a 2-dimensional input layer and an output layer working toward minimizing an objective function defined as the contextual information. The interconnection strength, composed by an internal state and a transient state with a non-linear self-feedback manner, is connected between neurons in input and output layers. To harness the chaotic dynamic and convergence process, an annealing strategy is also embedded into the ACCLN. In addition to retain the characteristics of the conventional neural units, the ACCLN displays a rich range of behavior reminiscent of that observed in neurons. Unlike the conventional neural network, the ACCLN has rich range and flexible dynamics, so that it can be expected to have higher ability of searching for globally optimal or near-optimum results.

Presents information on a study which proposed a cooling schedule in Hopfield neural network with annealing strategy to calculate the shortest path (SP) tree for multicast open shortest path first protocol. Shortest path computation with neural networks model; Annealing techniques; Description of the cooling schedule.

Segmentation of medical images, including Computerized Tomography (CT), Magnetic Resonance Imaging (MRI), mammographic imaging, and x-ray imaging, is an important step in the identification of tissue organization for analysis of the human body. Segmentation can also be cast as an optization problem that can be regarded as the minimization of a criterion function defined as a function of the Euclidean distance between two pixels with their local constraints. In this paper, the Hopfield neural network with penalized fuzzy c-means technique (called PFHNN) is proposed for medical image segmentation. Instead of using global information (such as histogram), the pixels with their first and second order moments constructed from their n nearest neighbors as a training vector are mapped to a two-dimensional Hopfield neural network for the purpose of classifying the image into suitable regions. The penalized fuzzy c-means technique is also applied to eliminate searching for weighting factors and to update training efficiency with additional bias input. In the experimental results, segmented data are obtained from test phantoms with added noise and MRI images.

We show an unsupervised parallel approach called the annealed Hopfield neural network (AHNN) with a new cooling schedule for vector quantization in image compression. The main purpose is to combine the characteristics of neural networks and annealing strategy so that online learning and hardware implementation for vector quantization are feasible. The idea is to cast a clustering problem as a minimization problem where the criterion for the optimum vector quantization is chosen as the minimization of the average distortion between training vectors and codevectors. Although the simulated annealing method can yield the global minimum, it is very time consuming with asymptotical iterations. In addition, to resolve the optimal problem using Hopfield or simulated annealing neural networks, the designer must determine the weighting factors to combine the penalty terms. The quality of the final result is very sensitive to these weighting factors, and feasible values for them are difficult to find. Using the AHNN for vector quantization eliminates the need for finding weighting factors in the energy function, which is formulated and based on a basic concept of the “within-class scatter matrix” principle. In addition, the rate of convergence is much faster than that of simulated annealing. The experimental results show that better and more promising solutions can be obtained using the AHNN than the LBG in image vector quantization. The convergence rate with various cooling schedules is discussed. © 1999 Society of Photo-Optical Instrumentation Engineers.

Hopfield neural networks are well known for cluster analysis with an unsupervised learning scheme. This class of networks is a set of heuristic procedures that suffers from several problems such as not guaranteed convergence and output depending on the sequence of input data. In this paper, a Compensated Fuzzy Hopfield Neural Network (CFHNN) is proposed which integrates a Compensated Fuzzy C-Means (CFCM) model into the learning scheme and updating strategies of the Hopfield neural network. The CFCM, modified from Penalized Fuzzy C-Means algorithm (PFCM), is embedded into a Hopfield net to avoid the NP-hard problem and to speed up the convergence rate for the clustering procedure. The proposed network also avoids determining values for the weighting factors in the energy function. In addition, its training scheme enables the network to learn more rapidly and more effectively than FCM and PFCM. In experimental results, the CFHNN method shows promising results in comparison with FCM and PFCM methods.

The kinetics and distribution of platelet deposition on protein-coated surfaces were studied in an in vitro flow system. When fluorescence-labeled platelets in whole blood flow through a flow chamber that is composed of a glass plate, platelets will deposit onto the protein-coated surface in a time-and location-dependent manner. A combination of fluorescence video microscopy and a digital image processing system allowed us to systematically study thrombosis kinetics under various flow conditions, with different biomaterials and forms of blood.In this paper, the image processing and pattern recognition techniques have been developed to quantify adhered platelets from numerous video frames. The dynamic adhesion status (newly attached, staying, detached) has been calculated automatically by software developed on an IBM/PC/386. These morphometric parameters as a function of flow rate and type of biomaterials can be determined experimentally.

Segmentation (tissue classification) of the medical images obtained from Magnetic resonance (MR) images is a primary step in most applications of computer vision to medical image analysis. This paper describes a penalized fuzzy competitive learning network designed to segment multispectral MR spin echo images. The proposed approach is a new unsupervised and winner-takes-all scheme based on a neural network using the penalized fuzzy clustering technique. Its implementation consists of the combination of a competitive learning network and penalized fuzzy clustering methods in order to make parallel implementation feasible. The penalized fuzzy competitive learning network could provide an acceptable result for medical image segmentation in parallel processing using the hardware implementation. The experimental results show that a promising solution can be obtained using the penalized fuzzy competitive learning neural network based on least squares criteria.