Search

To study the topographic distribution of the short posterior ciliary arteries (SPCA) entry sites into the choroid in normal eyes using structural en-face swept source optical coherence tomography (SS-OCT). Retrospective analysis of SS-OCT scans (wide-field structural SS-OCT 12 × 12 mm) of 13 healthy subjects was performed. Cross-sectional swept-source OCT scans derived from a volume scan were represented as en-face image display following the Choroid-Scleral Interface to obtain en-face OCT. SPCAs in their last scleral location before choroidal entrance were identified manually, counted and localized by two masked observers. Correlations between two masked observers were analyzed using inter- and intra-class correlation. Accuracy for the choroidal inner and outer border segmentation was 95–99%. Eighteen eyes from 13 normal subjects were included for SPCA analysis. The mean number of arteries was 13.8 ± 3.5 per eye. Thirty-six percent were in the center of the posterior pole image; however, 21% were in the temporal part of the posterior pole. Median accuracy of the detection is 0.94. The correlation between the two observers was fair (0.54). Our algorithm allows visualization of the SPCA at the posterior pole of the eye using wide-field en-face SS-OCT. It can also help the clinicians to study the SPCAs in numerous ocular diseases, particularly its relationship with focal choroidal diseases.

In this chapter, a cognitive computing model, a cognitive computing architecture, and a cognitive computing system are proposed and tested to address a big data classification problem. The proposed cognitive computing model is called the STE-M model and it adopts the standard components, senses (S), thoughts (T), experiences (E), and memory (M), of human cognition to describe the processes involved in cognitive computing for big data classification. Similarly, the proposed cognitive computing architecture is called the cognitive random forest and it amalgamates the STE-M model and a set of random forest classifiers to enhance continuous learning. It also includes intra- and intercognitive computing models to connect STE-M and random forest models and improve classification accuracy with spatial and temporal reasoning. A cognitive computing system is also proposed and it is used to validate the proposed cognitive computing architecture. Experiments with a robotic navigation scenario under different environmental conditions show that the proposed cognitive random forest is capable of handling the environmental conditions what we called the cune conditions for big data.

In this paper, a joint interference cancellation (IC) and decoding scheme is investigated for a multiple input multiple output (MIMO) space-time block coded orthogonal frequency division multiplexing (OFDM) based WLAN systems over frequency selective fading channels. A multiuser environment with K synchronous co-channel users, each equipped with two transmit antennas is considered. Assuming that the receiver is equipped with m (/spl ges/K) receive antennas, signals from K users are detected using minimum mean squared error (MMSE) interference cancellation followed by maximum likelihood (ML) decoding. It is shown that the joint IC and decoding scheme could be used to increase the system capacity (in terms of number of users or data rate) linearly with the number of co-channel users while maintaining the error rate performance. A theoretical error probability bound for the MMSE-IC scheme is derived.

In this paper, a computationally efficient adaptive channel estimation scheme is proposed for a space-time block coded (STBC) multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) system. This adaptive channel estimation scheme is based on recursive least square (RLS) algorithm. The proposed channel estimation scheme is less complex and bandwidth efficient than the conventional channel estimation schemes for MIMO-OFDM systems. Simulation results show that the proposed channel estimator performs quite close to the system where ideal channel state information is available. At the receiver, signals from different co-channel users are detected using minimum mean squared error (MMSK) interference cancellation (IC) method. It is shown that the signal detection with the proposed channel estimator could be used to increase the system capacity. Simulation results are compared with theoretical error bounds.

In this paper, an adaptive channel estimation scheme is proposed for space-time block coded orthogonal frequency division multiplexing (ST-OFDM) systems. In this scheme, a computationally efficient decoupled maximum likelihood (DEML) algorithm is used to estimate initial channel parameter values. Once the initial values are obtained, an adaptive algorithm is used to obtain more accurate channel estimate values. It is shown that the proposed channel estimation scheme can eliminate the error floor substantially and perform quite close to the system with perfect channel state information for nontime-selective fading channel. For time-selective fading channels, it has the ability to track the channel adaptively without training symbols.

This paper presents a space-time block coded multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) scheme over frequency selective fading channels. The system provides spectrally efficient transmissions to increase system capacity or system throughput for individual users. The proposed technique utilizes OFDM to transform frequency selective fading channels into multiple flat fading sub channels on which space-time block coding is applied. We consider a multi-user environment with K synchronous cochannel users, each equipped with n transmit antennas. The receiver is equipped with m(/spl ges/K) receive antennas. At the receiver, signals from K users are decoded using minimum mean squared error (MMSE) interference cancellation followed by maximum likelihood (ML) decoding. It is shown that the proposed scheme, with two cochannel terminals, doubles the data rate while maintaining the frame error rate (FER) performance as a single user scheme with some additional decoding complexity.

The two-dimensional discrete cosine transform (DCT) is probably the most popular transform in image and video coding systems as it achieves very high compression ratio for fast implementation. Various still digital image and video compression algorithms using this block-based transform coding method cause the artifacts, called the blocking, in the reconstructed images of a video sequence. This artifact is more prominent at low-bit rates. A new eight-neighbour symmetric post-filtering technique is introduced in this paper to reduce the widely known blocking effect which is commonly found in the reconstructed images of block-based transform coding schemes. In this technique, the weights of the eight neighbouring pixels of the pixel of interest are determined by the modified Lagrange's interpolation algorithm. The proposed technique reduces the blocking effect while maintaining the sharpness of the reconstructed image. The results obtained by the proposed technique are compared in this paper with the results obtained by several well known eight neighbour low-pass filters such as Reeve and Lim's method, the Laplacian eight neighbour accounting filter (LENAF), the Laplacian four neighbour filter with + form, the Laplacian four neighbour filter with x form and a two neighbour filtering method. These comparisons will be conducted with the following measures: peak signal-to-noise ratio (PSNR), block-edge impairment metric and subjective quality evaluation.

A quantitative distortion measure is proposed in this paper to measure the blocking artifact, which is a common problem in block-based coding mechanisms used in various digital image and video coding standards. The quantitative distortion measure, called Generalized Block-edge Impairment Metric (GBIM), has been improved by incorporating human perceptual characteristics based on local statistics and L/sub /spl infin// distortion measure along the block boundaries of a digitally compressed video image. This Improved Block-edge Impairment Metric (IBIM(L/sub /spl infin//)) gives results which are consistent and highly correlated with subjective evaluations.

This paper presents a new block classification algorithm for block filtering in the DCT-based digital video coding. In general, linear block filters are applied to every block on the reconstructed image regardless of their degree of visibility to reduce the blocking artifacts. The method proposed in this paper allows the coding scheme to identify the blocks that are apparent and significantly contribute to the overall blocking artifacts so that the block filters can be applied only to those blocks. The method calculates the block classification parameter from the original image and transmitted or stored to be used as a threshold value in the block classification process at the decoding stage.

Fast fuzzy filtering (FFF), based on fuzzy logic, is a new technique for detail-preserving smoothing of noisy image data. This paper shows how a hierarchical fuzzy approach can be used to perform detail sharpening of an image corrupted by noise. Results of computer simulation show that the proposed fuzzy approach performs better for further reduction of noise presented in an image.

A new technique based a genetic algorithm is proposed in this paper to find the optimal value of the ratio /spl Gamma/, which is a priori knowledge of the signal-to-noise ratio, in the Wiener filter for image restoration. The proposed method can be used in a control experiment in conjunction with any image-quality-assessment measure in order to obtain the optimal /spl Gamma/ for the measure. In the control experiment, the minimum mean square error (MSE) and image activity weighted error (IAWE) are used.

A method is proposed to measure the blocking artifact which is a common problem in block-based coding mechanisms used in various digital image and video coding standards. In this method, the quantitative distortion measure, the generalized block-edge impairment metric (GBIM), is modified to accommodate the local statistical characteristics, based on directional inter-pixel differences between adjoining blocks, of a video image. This modified block-edge impairment metric (MGBIM) gives results which are consistent and highly correlated with subjective evaluations.

This paper presents a new method for block classification, at the decoding stage, in digital image and video coding. Linear filters have been used to reduce the blocking artifacts caused by the block-based transforms used in the digital video processing. However, the linear filters have been applied to every block on the image regardless of their degree of visibility. In this paper, a block classification algorithm is proposed to identify the blocks that are apparent and significantly contribute to the overall blocking artifacts so that these blocks can be filtered out to reduce the blockiness.

Information visualization has been formalised as a continuous process of displaying information on computers. Information visualization systems used today are comprised of control and learning processes, however, they suffer from the incapability to automated job completion at the right time. Intelligent agents and multi-agent techniques have been employed to solve the real problem in which the adaptability and autonomy are essential. Most obvious application domains include database access via the Web, management of a computer network such as the Internet, electronic commerce, and industrial automation and control. This paper proposes an intelligent agents and a multi-agent system to model information visualization. The proposed techniques extend existing object-oriented representation techniques. The approach used provides a conceptual framework to guide system design and specification.

Due to rapid pace of technology advancement in the computer science discipline, both educators and students in universities have experienced difficulties in the teaching and learning of number of computer-oriented subjects. These include operating systems, computer organization and computer networking. Because of the nature of the material covered in these subjects, they must be taught using a practical approach. This practical approach can include various techniques used to present the course materials and concepts in a visual frame of reference. When using a visual approach to convey course material, both students and teachers encounter challenges. These challenges include greater monetary expense for required equipment, greater expense in terms of time to adapt the course material to a visual format regardless of the type of teaching methodology used (i.e. distance education teaching and learning). In this paper we propose a visualization technique to address a number of issues related to teaching and learning of an operating systems course in a university environment.

In this paper, an optimization technique using the Genetic Algorithm has been proposed to find the optimal value of the ratio Γ, which is a priori knowledge of the signal-to-noise ratio, in the Wiener filter to image restoration. The proposed method is suitable and effective for control experiments in conjunction with any image-quality-assessment measure in order to obtain the optimal Γ for the measure. It has been shown that the method finds the optimal Γ regardless of degree of noise degradation.

Over the last few years, streaming video and audio content over the Internet has become increasingly popular. Communication industries and individuals have successfully created various types of streaming media including live and pre-recorded television and radio programs, stereo-quality music, and various types of video and audio files. The purpose of this paper is to create a cost-effective virtual music video television station in which music and music video files can be sent from an audio/video station database through a communication line, such as the Internet, telecommunication channel and satellite transmission lines to the down-link processing station.

In this paper, the neuro-fuzzy technology is considered to model and control an adaptive information visualisation system. Three kinds of neural network used are: a general regression neural network (GRNN), a Kohonen's self-organized feature map (SOFM), and a fuzzy perceptron. The GRNN has been used to constantly predict the time of the redisplay dynamics of an adaptive information visualisation system, the Kohonen SOFM has been used for automatic generation of a set of initial membership functions and an initial fuzzy rule matrix, and the fuzzy perceptron has been used for controlling the redisplay dynamics of the system. The experimental results demonstrate the effectiveness and implementation of the proposed neuro-fuzzy control system. In particular, its performance of information redisplay has been improved up to 13% in terms of quantity of information per unit of time.

The soft computing techniques have been introduced in image processing in recent years and a number of promising results have been achieved. In this paper, two new techniques based on soft computing have been proposed to restore images that are degraded by blur and noise. The first technique is based on hierarchical fuzzy logic and the second is based on genetic algorithms. The hierarchical fuzzy approach is used to perform detail sharpening of an image corrupted by noise, whilst the genetic algorithm approach is used to find the optimal value of the noise-to-signal power ratio in the Wiener filter technique to restore blurred-noisy images. However, the second method can be used in a control experiment in conjunction with any image-quality-assessment measure. The measure used in this paper is the minimum mean square error (MSE) to demonstrate the effectiveness of the method.