Earticle

Fuzzy c-means algorithm is an unsupervised clustering algorithm, its clustering process can reduce the human intervention, and it is suitable for processing medical images of uncertainty and ambiguity. When simply using FCM algorithm in brain image segmentation will leads to the condition of low accuracy. On the basis of FCM algorithm, this paper proposes a new method which combines FCM algorithm and morphology algorithm. The result of simulation shows that this method can accurately and efficiently segment the brain image. The new algorithm is an effective method for image segmentation.

The analysis of MRI images is a manual process carried by experts which need to be automated to accurately classify the normal and abnormal images. We have proposed a reduced, three staged model having pre-processing, feature extraction and classification steps. In preprocessing the noise has been removed from grayscale images using a median filter, and then grayscale images have been converted to color (RGB) images. In feature extraction, red, green and blue channels from each channel of the RGB has been extracted because they are so much informative and easier to process. The first three color moments mean, variance, and skewness are calculated for each red, green and blue channel of images. The features extracted in the feature extraction stage are classified into normal and abnormal with K-Nearest Neighbors (k-NN). This method is applied to 100 images (70 normal, 30 abnormal). The proposed method gives 98.00% training and 95.00% test accuracy with datasets of normal images and 100% training and 90.00% test accuracy with abnormal images. The average computation time for each image was .06s.

Image inpainting is an important research topic in the field of image processing. The goal of inpainting is to recover the lost information of the target region from the rest of the image. Inpainting techniques can be applied in areas such as old photo restoration, object removal and demosaicing. Based on previous literatures of image inpainting and image modeling, we categorize inpainting processes of different methods and algorithms into the structure layer and the texture layer. Then the mathematical inpainting models and the formations of image impairment are analyzed and evaluated in detail. Experimental results are provided in the fifth section regarding to eight different algorithms measured by Peak Signal to Noise Ratio (PSNR) as well as direct visual perception.

Ultrasonic guided wave detection technology has numerous advantages compared with body waves. Accurately obtaining the time-of-arrival (TOA) or time-of-flight (TOF) of the ultrasonic guided wave signal is critical for the detection and location of defect because of the frequent use of the low frequency guided wave probe in testing. Thus, guaranteeing the propagation over a long distance and increasing the resolution of defect detection become the key points. This study adds the weighted nonlinear transformation functions to the minimum entropy deconvolution (MED). By adjusting the corresponding parameters, it can enhance a weak signal of small defects and suppress the noise signal. The experimental results indicate that this method can accurately obtain TOA to enhance the resolution of defect detection and can improve the speed of convergence effectively compared with MED.

Carbon nanotube is the new material which has ability to replace Si in the future. CNT has remarkable unique properties that make carbon nanotube a promising material in the future. Carbon nanotube field effect transistor is one of the main application of CNTs. Carbon nanotube Field Effect Transistor will play important role in designing of sequential and combinational circuits which are the base of digital computers. Carbon Nanotube Field Effect Transistors have been considered as accompaniment to , future electronic circuit due to the larger current carrier mobility in CNTs compared to bulk silicon. In this research paper, simulation of Low Pass Filter have done at 45 nm technology using hspice. The simulation result of proposed Carbon Nanotube Field Effect Transistor based Low Pass Filter show that the frequency response of Low Pass Filters are working satisfactory. It has applications in the low pass circuits. In electronics, these filters are widely used in many applications. Moreover, it is clear from the Phase response of Carbon Nanotube Field Effect Transistor based Low Pass Filter that it is stable Filter. So, we can use it in powerful conditions where stability is main concern.

GGAC is an improvement based on geodesic active contour model (GAC). GGAC model is a widely used method for image segmentation. But, it will be difficult to achieve satisfactory segmentation results to the texture, uneven structure, edge particles, weak edge and other features of the wood surface image. Therefore, the author proposes a segmentation method that integrates the improved Canny edge detection result integrated into the improved GGAC model redrawing boundary stop function, and uses the improved variational level set method to achieve the numerical solution. The algorithm has reduced the choice sensitivity to the initial contour and enhanced the scalability, which can make the profile curve converge to defect edges more rapidly, avoid the local optimum, and improve segmentation effects of weak edges and uneven image. The results are clearer, more consistent and real-time. It has provided a more effective way to segment the wood surface defects, and broadened the application scope of Canny operator and an improved geodesic active contour model.

It is assumed in the traditional total variation(neighbor domain, ND) algorithm that the pixel points are located at the edge and an edge-preserving model is set up. In the algorithm, pixels in flat regions of the image diffuse along the edge direction, leading to insufficient noise suppression and even presence of false edges. To carry on the edge-saving feature of neighbor domain algorithm and to make up its deficiency in omitting the image edge direction, this paper introduces direction neighborhood to the total variation algorithm so that edge points diffuse along the direction neighborhood. It changes the mode where edge points in the traditional ND algorithm diffuse along the multi-neighborhood, maximizing the smoothness along the edge direction and minimizing that at the vertical edge direction. The experimental results show that the image denoising method based on vector neighbor domain effectively addresses the drawbacks existing in the traditional ND algorithm and provides faster convergence efficiency, achieving both denoising and edge-preserving and improving PSNR and visual effects of the image.

A new facial feature position self-calibration method based on active computer vision is proposed in this paper to realize facial expression recognition. Compared with traditional method, the proposed method based on the extension focus thought only needs four linearly independent translational movements, one real rotational movement and one virtual rotational movement rather than the calibration reference object to realize the linear solutions orderly for internal reference matrix of camera, hand-eye relationship and feature point target depth. The experiment result shows that the proposed method is feasible and effective and the measurement errors of two-dimensional and three-dimensional feature points can be below 0.40mm, thus able to meet the industrial accuracy requirements.

To realize the gesture recognition of high precision ratio, the gesture recognition method of multi-model data fusion based on Kinect depth image is proposed, to implement the automatic splicing of models. First of all, the feature package model uses the speeded up robust feature (SURF) algorithm to replace the scale invariant feature transform (SIFT) algorithm to extract features, improve the real-time performance. Secondly, Hu moment is introduced to describe the global gesture features, further improving the recognition rate, the ray casting is used finally, and the obtained coordinate information is used to solve the rigid transformation between two point cloud models. Finally, the proposed data fusion method is verified through two experiments, the algorithm in this paper is better than the traditional support vector machine (SVM) method both in real time performance and recognition rate, and obtains better model splicing effect.

As the advancement in technology the number of application per chip is increasing and chip area remain same. If number of application per chip is increasing so the difficulty level of analog circuit becomes more and more complex. So there are various fault occur in analog VLSI circuits during manufacturing of any analog circuit. If these faults cannot diagnose and remove at initial stage then it will lead to various changes in output of system, which increase the overall cost. In this article we focused on modeling, analysis and diagnosis of various faults which occur in analog VLSI circuit. We describe new approach to diagnose parametric and catastrophic fault in analog circuits with the help of signal flow graph technique. This technique is very simple and structural in nature. It is applicable to various linear analog VLSI circuits. In this paper we implement this approach on MIMO (Multi Input Multi Output Circuit).All the equation and model for MIMO circuit and simulation are done with the help of MATLAB/Simulink tool.

Immature green citrus fruit detection using conventional color images is a challenging task due to fruit color similarity with the background, partial occlusion, varying illumination and shape irregularity. Therefore, most existing green fruits detection algorithms, which use color as the main discriminant feature, have a low recognition rate and a high rate of false positives. In this manuscript, we developed a novel Green Citrus fruit Detection algorithm based on the proposed Reticulate Grayladder Feature (GCDRGF), which contained 4 major steps: First, an 8-graylevel image was generated by the preprocessing steps of median filtering, histogram-based equalization and 8-graylevel discretization of the input raw image. Secondly, reticulate grayladders were obtained by a multidirectional scanning on the 8-graylevel image, and rule-based pseudo-grayladder removal strategies were used to remove false positives of target grayladders. Thirdly, grayladder clustering and fruit location fitting were used to generate candidate regions for target fruits. Finally, majority voting was performed to determine the results of candidate regions based on the analysis of apparent features and recticulate grayladders within candidate regions. The experimental results proved the effectiveness of the proposed reticulate grayladder feature and the corresponding detection algorithm with respect to various illuminant and imaging conditions. Compared with the existed eigenfruit algorithm, our algorithm has a higher rate of successful recognition and a lower rate of false positives, which helps to greatly improve the productivity of robotic operations.

To lower the mapping complexity of designing analog decoders, a method to optimize the design of low-density parity-check (LDPC) analog decoders is proposed in this paper. Based on factor graphs and the sum-product algorithm, the LDPC decoding process on the factor graph and the construction of analog decoders are exploited. Then the frequent subgraph mining algorithm is introduced to search the isomorphic subgraphs in factor graphs. According to the output of the frequent subgraph mining algorithm which enumerates all the subgraphs in factor graphs, the mapping complexity of a LDPC analog decoder can be significantly reduced. Finally, a (40, 16) LDPC analog decoder is constructed using the proposed method. Simulation results show that the need to place gates and connections can be reduced 90% and 23%, respectively, and the ideal performance is obtained by carefully choosing unit currents and decoding time.

The sensing matrix has an important influence on the original signal sampling and reconstruction algorithm in the compressed sensing theory. A complete random sensing matrix has the drawbacks of large storage and high complexity in its implementation. In this paper, we propose an interlaced filling algorithm to construct the sensing matrix, which has a quasi-cyclic structure for efficient hardware implementation. The new sensing matrix has small coherence, which provides assurance for the recovery of sparse signal. Meanwhile, some experimental comparison with the other sensing matrix is accomplished. The simulation results demonstrate that the proposed sensing matrix not only obtains better performance but also owns easy hardware implementation.

For the situation with unknown qualities of local tracks in sensor networks, a new tracklet-global track fusion method using the support degree function (SDF-T2GTF) is proposed. According to the characteristic of actual transmission modes, two local estimates of a moving target in adjacent interval transmitted by the same local node are defined as a tracklet, and subsequently tracklet-global track (T2GT) fusion can replace the traditional track fusion in the global node, namely local track-global track (LT2GT) fusion. Considered the advantage of the fuzzy track association (TA) method for unknown prior information of local tracks, it is used in T2GT association. Then all correlated tracklets in the same interval can be mapped into a set of points in parameter space by the Hough transform (HT) algorithm. The support degree function of these points is utilized to dynamically estimate the qualities of tracklets and reasonably allocates the weights of local estimates in fusion results. Hence, the proposed method can realize T2GT fusion without the prior information of local tracks. The experimental result illustrates that the proposed method can satisfy the requirement of data transmission in real systems, and can realize T2GT fusion; on the other, it can improve the performance of track fusion in accuracy compared with the traditional methods.

Speech processing involves making changes to speech signals for a required application. Therefore it is very important in the area of communications. In the 21st century the use of FPGAs has been more prevalent in many areas. In this paper, the impact of FPGAs on speech processing is evaluated. Speech processing algorithms are classified and the current standards are overviewed. The role of micro-processors in speech processing is evaluated. Current implementations of FPGAs in speech processing are outlined. Finally, aspeech processing application that does not use FPGAs is evaluated on the basis that itis implemented using FPGAs.

It is significant that adopt text recognition at mobile devices to care human’s health. We observed that although OCR is very suit for recognizing scanned documents, it has poor performation on mobile photoes, which suffer from unequal lighting, clutter, skew, or poor image quality. Therefore, a new algorithm is proposed that take a series of measures to deal with these tough situations of mobile images. This work includes three main steps. Firstly we adopt perspective correction to rectify the distortion of an image. Secondly we use filter to further eliminate the effect of noisy in image. Finally we apply text segmentation to effective measure each text row of image. Compared to OCR text recogniztion success rate 34.7%, the success rate of our method is 65.8%. Experimental results show that the proposed algorithm greatly improves the accuracy of text recognition.

The general edge extraction algorithm is not ideal to process infrared images, which is low contrast and blurred edge. In this paper, we used the multi-fractal spectrum to edge of infrared image. We extracted the edge information of the image and calculate the measure and fractal spectrum with multiple singular values of each pixel. Analysis of the similarities and differences of multiple measure, the function in edge extraction, meanwhile, analyzed the fiction of fractal characteristics to edge image extraction. This method differs from the traditional gradient algorithm，It determines whether the edge or not just according to the local extreme points , but according to the pixels in the local and global relationships to determine whether the pixel is a real edge. It can neglect important edge pixel and texture pixel，which is more in line with the human visual mental. It provide a good reference for recognition of infrared image and further processing.

Optical Wireless Communication (OWC) also known as Free Space Optics (FSO) technology has many advantages such as availability of license-free spectrum, low deployment cost, large modulation bandwidth, low consumption of power, and small size. In this paper, the investigation on the performance of free space optical communication link has been done for different system parameters and transmission windows using OOK (on-off keying) modulation scheme. The performance of the system has been analyzed using Q Factor and SNR of the received signal as the performance metric. Also, the effect of atmospheric attenuation on the performance of the system has been analyzed by varying the link distance between the transmitter and the receiver for a specified transmission power level and bit rate. The performance of the system has also been investigated under different data transmission rates and transmission power levels using OPTISYSTEM simulation software.

Recently, the compressive tracking (CT) method has attracted much attention due to its high efficiency. However, the CT extracts samples around the previous target region within a fixed search radius; the searching area is unsuitable when the target undergoes abrupt acceleration change. Meanwhile, the classifier learns the features of the target online without judgment even the target is fully occluded. Thus, the improper searching area and incorrectly updated features lead to a marked drop in precision of tracking. To solve this issue, a robust target tracking method integrating spatio-temporal model to constrain the searching area is proposed in this paper. Different from CT, the proposed method initially constructs the spatio-temporal model to calculate a confidence map between consecutive frames, and the region with high confidence suggests the high possibility that target exists. Thus the samples can be extracted in the high confidence area. Then, the optimal target location can be estimated with a naive Bayes classifier using sparse coding features. Experiments show that the proposed method outperforms several competing methods in efficiency and robustness.

Human-computer interaction (HCI) has great interactive applications in many fields. A novel eye-hand coordination based HCI approach has been proposed. According to the fact that the eye gazing starts prior to the hand movement and reaches the target in advance, an eye-hand coordination model in a non-contact way is constructed by facial orientation and skeletal joints of hand. Both temporal median filtering and moving average filter strategies are developed to overcome some fluctuation influences during HCI. The diversity of interactive habits among multiple users is considered in an ordinary hardware from a crowded scene without any hypothesis for the scenario contents in advance. Comparative comparisons with state-of-the-arts have highlighted the superior performance of the proposed approach.

This Paper serve as investigation for the different behaviors of resonant modes of the metamaterial loaded shorted Microstrip circular Patch antenna where the inner core is loaded with MNG(μ negative) metamaterial. The material DPS is outside of the MNG Core material. DPS is a double positive regular dielectric. MNG metamaterial loaded with circular patch is used for frequency tuning and dual band operation. Simulation has been performed by way of CST software program.

With the rapid advancement of the computer assisted medical applications, the MRI image segmentation has been a hottest research topic. In the neural network is used for the image segmentation, we need a lot of training data, because of the large amount of the data, computing speed is quite slow, not suitable for real-time data processing, lead to the low resolution image segmentation, the resolution is not high, this paper proposes a fuzzy image segmentation algorithm of the BP neural network. Fuzzy set theory is used to subtract the characteristics after area of the image segmentation, reduce the dimension of feature vector. We adopt the revised neural network to undertake the experimental simulation compared with the other state-of-the-art approaches. The result proves the effectiveness of our methodology. Our algorithm could segment the regions of interest with the ability of eliminating the out side noise which achieves the better robustness.

A face recognition algorithm based on local binary Haar feathers which represented as Kadane optimizing multi-threshold AdaBoost was proposed according to the problems of texture shape feature representation and classification algorithm accuracy in the process of facial classifying detection and recognition, First, improve the traditional expression by using image Local binary pattern of Haar features , improve image model of texture and shape feature expression ability ; Secondly, for single threshold weak learning algorithm we can not make full use of local binary Haar feature information, resulting in a lower classification accuracy problem proposed Kadane optimizing multi-threshold AdaBoost classifier, to achieve local binary Haar feature representation of facial high accuracy recognition; Finally, through the experiments show, efficient face recognition rate can reach more than 90% by the algorithm,which is superior to the selected comparison algorithm.

Nowadays, as various image manipulation tools are available very easily. Any person having a little knowledge about these tools can doctor the available images. So digital images are no longer trusted. Computer graphics and digital photography have made the tampering over image easy to commit but hard to detect. Although various image forgery techniques are available but copy-move image forgery is one of the most hard to detect image forgery. In Copy-Move image forgery a segment from the original image is copied and after performing some manipulation over that, segment is pasted at some other location on the same image. This forgery is intended to hide noticeable information shown by the image or for adding information in original image to convey a wrong message. We cannot identify such forgery on the basis of incompatibilities present in an image because the copied segment is taken from the same image so the properties like noise, blur, texture, color palette remain similar to the original image. So, copy-move image forgery is a serious threat to Image forensic Investigators. Researchers have developed several methods for detecting such kind of forgery based on exhaustive search and block based methods. Block based method is more successful in detecting such kind of forgery due to its speed and less complexity. In this paper we discuss forgery detection techniques based on Discrete Cosine Transform and Discrete Wavelet Transform.

Image fusion is an important branch of information fusion, which is widely used in various fields. At present, the image fusion method is mainly aimed at the different frequency information of the images, the images are fused in transform domain. But in practical application, image fusion is used to improve the credibility of the target information and the demand of background information of is not high. Therefore, this paper puts forward an image fusion method combining with image classification. Firstly, the NSCT transform is used to transform the source images, and the K-Means method is used to realize the classification of the target and the background, and the different fusion criteria are used to get the target and the background. The experimental results show that the image fusion based classification method has a better effect on the subjective visual effect and objective evaluation index.

In order to improve watermark capacity, image quality and error correction capability of DFT-based watermark algorithm, this paper proposes a novel efficient image watermark algorithm based on DFT transform. Firstly, the watermark information is encoded by using hybrid error correcting code, and the watermark is embedded into an image, and the code ensures that the correct rate of watermark extraction is 100% in the condition of the watermarked image under attacks. Secondly, through changing the way of embedding watermark into the frequency domain data block, watermark capacity can be increased to 2 times of the original algorithm. Thirdly, through extending the frequency domain region of embedding watermark, visual effect of the watermarked image is obviously improved and watermark capacity can be increased. The experimental results show that the proposed algorithm is significantly better than the original algorithm in robustness, and it can ensure that the correct rate of watermark extraction is 100% after the watermarked image suffered with attacks, and the proposed algorithm and the original algorithm are basically the same in PSNR, and the proposed algorithm is better than the original algorithm in visual effect of watermarked image.

The face describes the personality of humans and has adequate importance in the identification and verification process. The human face provides, information as age, gender, face expression and ethnicity. Research has been carried out in the area of face detection, identification, verification, and gender classification to correctly identify humans. The focus of this paper is on gender classification, for which various methods have been formulated based on the measurements of face features. An efficient technique of gender classification helps in accurate identification of a person as male or female and also enhances the performance of other applications like Computer-User Interface, Investigation, Monitoring, Business Profiling and Human Computer Interaction (HCI). In this paper, the most prominent gender classification techniques have been evaluated in terms of their strengths and limitations.

Along with the advancement of Internet and digital technologies, more and more Chinese traditional paintings are becoming available on the Internet. As a result, computerized indexing and classification of given Chinese paintings emerge to be one of the focused research areas over recent years. As traditional Chinese paintings rely on the special drawing tools to illustrate the artistic styles, it distinguishes from western paintings in terms of strokes, contours, color and textures etc. Additionally, drawing lines play important roles in most traditional Chinese paintings. Consequently, the existing research on Western paintings is normally not applicable to traditional Chinese paintings. In addition, color-based approaches are also not applicable as traditional Chinese paintings mostly rely on gray scale texture to express their art styles and content. This thesis reports my intensive research program on computerized classification and automated learning and analyzing techniques for traditional Chinese paintings, in which a number of novel research and ideas are developed and put forward for style-based classification as well as its related theories and new concept introductions. My own novel contribution can be highlighted in machine learning, especially one-class SVM (OCSVM) based classification of traditional Chinese paintings. In this paper, the one-class SVM technology is revised to introduce a supervised learning element and arranged into a parallel OCSVM classifier. Based on the statistics features, a new concept of enforced learning has been introduced to remove the false positives at each learning cycle together with a new scheme of adaptive upgrading of decision parameters. Extensive experimental results support that the proposed new classifier achieves significant improvements in comparison with existing representative techniques.

Image classification is an important task in computer vision. The methods based on spatial information generally employ some low-level features for image classification, such as gray scale, color, texture and location. It is difficult for vision system to understand and the single feature is too limited to obtain correct classification results. In this paper, an algorithm based on multi-kernel feature learning is proposed and used for image classification. First, the kernel function is used to produce a kernel descriptor, which aggregates the pixel attributes into patch-level features; Then, through the multi-kernel learning, these descriptors are further aggregated to obtain hierarchical multi-feature descriptors; Finally, the label of each image is given by the fusion strategy of on multi-classifiers, which effectively utilizes the advantages of multi-kernel learning and takes the complementary among the classifiers into account. The experimental results show that the proposed method is efficient in promoting the classification results.

The periodic estimation method of the PN sequence in DS Signal is researched. It is hard for traditional methods to estimate PN period of DS/SS accurately in the condition of low Signal-to-noise ratio. On the basis of the traditional spectrum reprocessing and the cepstrum method, a PN period estimation method which has integrated serial average with spectrum reprocessing and the cepstrum are put forward to solve this problem and meet the requirements of lower Signal-to-noise ratio. The simulation results show that, when the period of PN sequence is estimated accurately, the cepstrum method has better estimating performance than the spectrum reprocessing method, and the spectral estimation method has lower Signal-to-noise ratio tolerance after improvements.