Earticle

Image denoising is a basic problem in image processing, due to the image structure has the characteristic of self-similar, using the ideas of nonlocal, this paper proposes a non-local denoising method based on sparse representation, the structure information of image is improved after denoising, at the same time making similar image tiles have similar sparse representation, image reconstruction effect is better, through the numerical simulation the results show that the method has good application value.

In this paper we have proposed a new marker selection technique using Support Vector Machine over-fitting. Markers are the most reliable pixels in a class. We used our proposed technique to do classification of hyperspectral image with very low training samples, as low as one pixel per class. We have used both spectral and spatial information to improve the classification results. The spatial information is extracted using Extended Morphological Profiles with duality. Nonparametric supervised feature extraction methods are used to eliminate the redundant and irrelevant information in both spatial and spectral domains. In the end we have done experimentation to verify our proposed approach. The experimentation results show that when non-parametric weighted feature extraction method is used we get better classification results. The classification maps shows that even with just one training sample per class we still can get a reliably reasonable classification map.

As a kind of very effective methods of gathering data processing - principal component method, it can be used to determine the variables between the linear combination rule, reduce the dimension of feature space, select the optimal variables instead of the original. In recent years, as in the field of image processing, a wide range of application of principal component inspection technology, its shortcomings are also needless to say, the main components of investigation can only in the presence of one dimensional vector. Plane principal component, but can be on the premise of reducing data transformation between time, directly with two-dimensional vector matrix, which results in better image processing speed operation. On this basis, in the light of the characteristics of the remote sensing images, principal component and on the plane algorithm combining wavelet transform, put forward a kind of based on wavelet transform and principal component of the denoising algorithm. Experimental results show that the proposed method is better than first when some typical denoising method, this method can effectively remove gaussian noise of remote sensing images, made in the image edge details such as information can be more perfect.

In modern life, we see more techniques of biometric features recognition have been used to our surrounding life, especially the applications in telephones and laptops. These biometric recognition techniques contain face recognition, fingerprint recognition and iris recognition. Our work focuses on the face recognition problem and uses a deep learning method, convolutional neural network, to solve it. And we use the Sobel operator to improve our result accuracy. LFW dataset is used for training and testing which gets a considerable result. And we also test our system on other face dataset, which also has a high accuracy on the recognition.

In this paper, an effort has been made to observe critically the impact of various channel coding and signal detection schemes in DCT-IFDMA aided SC-FDMA wireless communication system on color image transmission. The simulated system incorporates MMSE-SIC , OSIC, ZF and MMSE as signal detection and ½ -rated irregular LDPC and Repeat and Accumulate as channel coding schemes under 4 × 4 antenna configuration. It is noticeable from MATLAB based simulative study that the system shows quite satisfactory performance in retrieving transmitted color image under scenario of hostile fading channel environment with implementation of MMSE-SIC signal detection, QAM digital modulation and Repeat and Accumulate channel coding scheme.

With the development of image processing and network technologies, more and more video applications emerged. To ensure quality of service for simultaneously transmitted multimedia and data streams, DiffServ (Differentiated services) is proposed. However, the current standard of DiffServ is not suitable for video streaming. In this paper we propose a cross-class priority (CCP) based video streaming scheme, in which packets of different frame types are assigned to different traffic classes to occupy more scheduling opportunities. Simulation results show that more video packets could be received when CCP policy is adopted. Thus CCP policy can tolerate a relatively high data rate of data stream. Scheduling mode and video sequence employed, and settings of RED parameters of different traffic classes are factors that influence the performance of CCP based scheme too.

Segmentation is a efficient technique of dividing the image into different regions or segments. Most of the researchers took clustering as the best method of segmenting an image. In clustering we try to increase the similarity within a same class and decrease the similarity between the classes. Many clustering algorithms were developed like FCM, FLICM and FELICM which are considered as the best algorithms to cluster the data. In our paper, we combine FELICM (Fuzzy Edge and Local Information C-Mean) with the negative selection algorithm. Negative selection algorithm is an evolutionary method which is based on artificial immune systems. The proposed method result shows us high accuracy results and even solves the problem of over segmentation.

Local Graph Structure (LGS) and its variation Symmetric Local Graph Structure (SLGS) have been proven to be effective for image recognition. However, they have shortcomings without considering the contribution of the difference between the target pixel and the surrounding pixels, and the difference between surrounding pixels to the feature value of the target pixel. To overcome the shortcomings of the traditional methods, this paper proposes a Difference Symmetric Local Graph Structure (DSLGS) algorithm for finger vein recognition. The DSLGS operator considers the contribution of different value to the feature of target pixel, making the extracted feature more stable. The experiment results show that the proposed algorithm has better performances t

The current in frequency domain can not be analyzed at the same time because of the limitations of the Fourier transform algorithm. Hence, the wavelet transform method is used to analyze the short circuit transient current and calculate the negative-sequence component in this paper. Then, the non-periodic component and periodic component fundamental wave of the transient current are obtained, and the high harmonic components and decreased periodic component are removed. The non-periodic component is seriously considered in the transient negative-sequence component calculation. The calculation results indicate that the transient negative-sequence component contains two parts, the equivalent negative-sequence component of non-periodic component and the negative-sequence component of negative-sequence current. The world's first AP1000 third generation 1250 MVA nuclear half-speed (4-pole) turbo-generator is adopted as the calculation model. The transient parameters of the large generator are calculated on the principle of transformed current waveform in the internal short circuit fault condition. Furthermore, the transient negative-sequence component of the large-capacity generator is also calculated and the negative-sequence ability is deeply discussed in this work, which provides a theoretical basis for the large generators’ protection and design.

Due to powerful computers and advanced photo-editing software tools the manipulation of images has become an easy task. Confirming the authenticity of images and detecting tampered regions in an image without any knowledge about the image content is an important part of the research field. An effort is made to survey the recent advancements being made in the field of digital image forgery detection and thus passive methods for forgery detection are being presented. Blind or passive methods do not require any explicit former information about the image. In the first part, various image forgery detection techniques are classified and then an overview of passive image authentication is presented and the existing blind forgery detection techniques are reviewed.

In this paper, we propose an image interpolation algorithm based on the characteristics of the Human Visual System (HVS). Firstly, the saliency map is extracted for the input image by the feature contrast. Based on the saliency detection algorithm, the input image can be divided into salient regions and non-salient regions. A non-local means based interpolation and the bicubic interpolation operations are adopted to implement the image interpolation for salient and non-salient regions in visual scenes, respectively. Experimental results show that the proposed method improves the perceptual quality of the salient regions during image interpolation. The PSNR and SSIM results demonstrate that the proposed method can obtain better performance than existing related image interpolation methods.

Microwave ferrites require high coercivity for their use as microwave absorbers. Strontium hexaferrite is one of the best material due to its magnetic properties , high coercivity, large unilateral magnetic anisotropy and very low cost in comparison to others make it suitable for the use as microwave components specifically microwave absorber .Strontium nano hexa ferrites were successfully synthesized through chemical co-precipitation method. The synthesized nanostructure was characterized by X-Ray powder diffraction(XRD) and fourier transform infrared spectroscopy (FTIR).The XRD result shows the shift from amorphous to crystalline after the calcination of sample at 6000C.The behaviour of ferrite as microwave absorber was also studied for X band of frequencies.

This paper presents a forecast model of fuzzy time series prediction to study the prediction of enrollment. It takes the percentage of enrollment changes as the domain, constructs the inverse fuzzy number and predicts the enrollment of Qiong Zhou University from 2005 to 2013. Compared with the existing models, the mean square error and prediction error of the ameliorated model are smaller and the precision is higher. The authors get the better method to solve the prediction problem based on the inverse fuzzy prediction model.

A digital watermark is a type of indicator covertly embedded in the noise tolerant signal such as an image or audio data. Digital watermarking is the method that embeds information known as a watermark into the multimedia object such that the watermark can be extracted or detected later to create an assertion about the object. Embedded watermark will permit recognizing the owner of the work hardware implementation. This idea is applicable for the digital video and audio also. Embedding a digital signal (image, video or video) with data which cannot be easily eliminated is called digital watermarking. Digital watermarks may be used to the authenticate the integrity or authenticity of the carrier signal or to display the identity of its owners Digital watermarking is used as a key result to create the document transferring protected from unlawful interferences. Digital watermark methods are used in numerous areas such as copyright owner identification, protection and broadcast monitoring. Digital watermarking is the processing of combined information into a digital signal. A watermark is a secondary image, which is overlaid on the host image, and provides a means of protecting the image. In this paper, aim is to present a survey of numerous methods on the basis of digital image watermarking. The digital watermarking method is becoming more important in this developing society of internet. The goal of this implementation is to survey the different techniques for Digital Watermarking.

For the problem of electrical energy measurement could not be united under the condition of non-stationary distortion signals, in the paper, the functionals series was used to modeling the signals. Firstly, the Wiener kernel of nonlinear load was acquired; the output signal of nonlinear load was expressed using functionals series under the condition of sinusoidal excitation; and then, we used the wavelet transform to processed and acquired component of fundamental and distortion; and though the acquisition component of voltage and current, we analyzed the power flow of nonlinear load, and according the result of power flow, we proposed a new method of active power measurement. The simulation result shows that the result of theoretical analysis is the same as the new method, and the new method of electric energy measurement based on the functional series can achieve the reasonable electrical energy measurement.

Due to the normal forecasting methods for subgrade settlement using observation data have different explicabilities, and the predicting results has bigger volatility and lower accuracy. The Combined forecasting model of subgrade settlement based on forecasting availability and real-coded quantum evolutionary algorithm (RQEA) is put forward in this paper. At the first, according to the basic settlement law of subgrade and characteristics of settlement curve, the growth curve with the S-type characteristics are chosen as single forecasting model; Then, to get the weights of each single forecasting model, objective function is build on the basis of standard of forecasting availability maximization, and RQEA is employed to solve the objective function, and to construct the combined forecasting model of subgrade settlement. The result of engineering practice shows that the proposed method has better prediction accuracy and stability, and can meet engineering demand.

In image processing, computer vision and pattern recognition, the Image retrieval is a most popular research area. In this paper, performance of various CBIR systems, based on combined feature i.e color texture and shape, are compared.

A new finger vein recognition method based on two-dimensional principal component analysis (2DPCA) and kernel maximum between-class margin criterion (KMMC) is developed. The algorithm includes four stages. Firstly, perform preprocessing steps which include normalizing and mean-filtering on the finger vein images, secondly, employ the 2DPCA to condense the dimension of image vector, thirdly, apply the KMMC to reduce the dimension of training samples further, and finally, take the match and recognition step by computing the Euclidean distance between each sample. Our experimental results indicate that the new method has good recognition effect.

Hiding the data in digital images has been area of interest in the digital image processing domain. Although so much work has been carried out in the literature to resolve the issues like increasing the data capacity, creating the secret image alike of target image but most of the works fails to meet the practical requirements. This paper presents an approach where mosaic image generation has done by dividing the secret image into fragments and transforming their respective colour characteristics into corresponding blocks of the target image. Usage of the Pixel colour transformations helps to yield the lossless recovered image based on the untransformed colour space values. Generation of the key plays an important role to recover the data from the secret image in lossless manner. Finally the same approach can be performed on videos also which helps to eliminate the flickering artifact to achieve the lossless data recovery in motion related videos. The experimental results shows good robust behavior against all incidental and accidental attacks and compare to the conventional algorithms performance evaluation has been increased in a significant way. The result of the proposed technique outperforms present techniques and the results are simulated using MATLAB.

According to existing methods of singularity detection in stress wave signal, raise and define the quantitative information entropy and the mean vector model of Information entropy, the definitions are based on probability theory and mathematical statistical theory and information entropy, Define the method of singularity detection based on the quantitative information entropy. This method is tested and simulated using stress wave of pile detection and compared with the modulus maxima method. The experimental results show that the quantitative information entropy method has advantages in anti-noise performance and can achieve a high accuracy and locations information compared with the modulus maxima method. In the quantitative information entropy method, multidimensional signal is processed by dimensionality reduction and can be widely applied to various research fields, such as fingerprint identification and facial identification.

By using MATLAB, a kind of software decoder is designed which has filter and fault correction function and is used in aviation radio communication. Through Morse code text, using dits and dashes to uniquely represent characters such as alphabets、numbers and punctuations (41), those characters are coded through algorithm. A new decoding algorithm is designed which is composed of filtration, binary, difference and zero removal operation. This decoding method is robust and interference immunity, it also has such functions as filtering and fault correction. It can filter the random noise and Gaussian noise so as to reduce the error rate of aviation radio communication and thereby increase the quality of aviation radio communication. This method is checked though an example to decode the word “Hello”.

Paddy is the staple food of India and many other countries. It is very essential to find out the best variety that promises good yield. This paper presents a methodology to identify variety of paddy field images. In this work, we have considered 22 varieties of paddy field images and they are divided into three classes based on physical features as light green, lush green and pale green. Identification is done using color, texture and combination of both types of features. Color features are extracted using HSV and texture using GLCM. Artificial Neural network (ANN) is used for identification of variety of paddy field images. Considering only color and texture, the results were not satisfactory. Combined features resulted in n accuracy of 85.7% in light green, 83.1% in lush green and 100% in pale green class. The work is an attempt to disseminate knowledge about new variety of paddy crop required to promote the large scale cultivation.

In this paper, we present for printed multi-oriented, multi-scaled and noisy Greek characters recognition a comparison in terms of precision, rapidity and stability between several classifiers which the first one is a probabilistic that is hidden Markov model, the second is a neuronal that is Kohonen network or self-organizing maps while the rest of other classifiers are based on a combination between these both classifiers and even more a statistical method that is K nearest neighbors in their tree different versions which are majority voting, weighted distances and fuzzy. For this purpose we have for pre-processed each character image by the median filter and the thresholding technique, then in order to extract efficiently their features, we have exploited the Krawtchouk invariant moments.

Aiming at the complexity of ground objects in urban area, and the difficulty in distinguishing ground objects using spectral characteristics, we extracted normalized different indexes, namely Modified Normalized Difference Water Index (MNDWI), Soil Adjusted Vegetation Index (SAVI ) and Normalized Difference Building Index (NDBI) , as the key auxiliary information for land use classification of urban area. To solve problems of RBF neural network, such as local minimum values and discrete output value in output layer, we used max-min distance means to initialize RBF center, and introduced equilibrium factor into Gauss function to improve RBF neural network learning algorithm. On this basis, a new urban area classification model was proposed based on improved RBF network and normalized difference indexes. At last, NanChong city in SiChuan province of China was taken as the study area, and TM images was used as experiment data to test the model proposed in this paper. The results showed that, based on the improved RBF network, with the help of spectral band information, the classification overall accuracy was 89.97%, Kappa coefficient was 0.88; using both spectral band information and normalized difference indexes, the classification overall accuracy was 95.02%, Kappa coefficient is 0.94, the classification overall accuracy was improved by 5.05%. Also, the experiment results showed that, with the help of spectral band information and normalized difference indexes, the classification overall accuracy of MLC, BP and improved RBF network was 90.12%, 93.63%, 95.02%, respectively, which means RBF has an advantage of fusing geological parameters in classification.

In the recent years, dehazing images have been extensively studied by researchers in various applications like in traffic monitoring, video surveillance, video security, marine surveillance etc. Various methods that make use of single image have been proposed such as: Dark Channel Prior (DCP), Improved Dark Channel Prior (IDCP), IDCP with Guided filter, Anisotropic Diffusion and DCP with Histogram specification. This paper is an effort to compare the above mentioned techniques on the basis of picture quality and parameters like Contrast gain (Cgain), Color Naturalness Index (CNI), Number of saturated pixel (σ), Normalized Color Difference (NCD) and Time Complexity (TC). It is observed that the best perceptual quality is obtained for IDCP with Guided Filter followed by IDCP, DCP with Histogram Specification, Anisotropic Diffusion and DCP.

As the wheeled mobile robot is widely used in various fields, requirements of control accuracy for wheeled mobile robot are also increasing. Vision sensors get more and more attention because they are information capacity, high efficiency, non-contact measurement. The servo control problem of robot visual has also become a research hot spot. Dividing from the number of vision sensors, visual servo system can be divided into monocular visual servo system, binocular visual servo system and multi-purpose visual servo system.