Earticle

The primary challenge to design a mobile sign language communication system is overcoming the limitation of battery power. A scheme which allocates the computational resource of the encoder adaptive to available battery power and deaf people’s visual system is proposed. In the scheme, encoding levels which determine number of reference frames and search range are adaptively selected according to the battery power and frame complexity at frame level. Then possible partition mode and quantization parameter are adaptively adjusted at the macro block (MB) level according to the relative priority of each MB. Experimental results show that the proposed algorithm obtains better peak-signal-noise-rate of face and hands that improves the intelligibility of sign language video. Encoding time can be saved by up to 89.4% compared to the standard H.264 encoder. Increased bit rate of the proposed algorithm is less than 5%, which is negligible.

New innovative research trends are more essential in the area of image compression for various imaging applications. These applications require good visual quality in processing. In general the tradeoff between compression efficiency and picture quality is the most important parameter to validate the work. The existing algorithms for still image compression were developed by considering the compression efficiency parameter by giving least importance to the visual quality in processing. Hence, we proposed a novel lossless image compression algorithm which efficiently suited for various types of digital images. Thus, in this work, we specifically address the following problem that is to maintain the compression ratio for better visual quality in the reconstruction and considerable gain in the values of Peak Signal to Noise Ratio (PSNR). We considered medical images , satellite extracted images and natural images for the inspection and proposed a novel procedure named as Novel Optimized Golomb Rice Coding (NOGR) to increase the visual quality of the reconstructed image. The result of the proposed technique outperforms present techniques and the results are simulated using MATLAB.

Most of the noise in speech communication lines can be considered as Gaussian white noise. Voice activity detection (VAD) in noisy environment is an important process in many speech signal processing algorithms. Unlike the other VAD algorithms, this paper proposes a simple and novel VAD algorithm based on the probability distribution function (PDF) of FFT magnitudes of both clean speech and Gaussian white noise. When the signal-to-noise ratio (SNR) is high enough, the method using Gamma distribution to detect the speech performs well, while the method using Rayleigh distribution under lower SNR can be complementary. In addition, the threshold to determine which method to use is presented based on the tests under different SNR. Simulation results show that the proposed algorithm is efficient. Both the hardware and software of a low cost system for VAD are introduced, with the proposed algorithm achieved in a digital signal processor (DSP). Each detection takes on less than 100 ms, which can be used for real-time processing.

Block-matching and 3D filtering (BM3D) denoising algorithm [1] proposed recently has a problem of computational burden especially for low noise level and a sharp performance drop for high noise level. To solve it, an improved version of BM3D is proposed. The solution combines the digital image characteristic with added noise pollution levels, and adaptively selects block-matching threshold in grouping stage. Experimental results demonstrate it outperforms not only in terms of objective criteria of PSNR and running time, but also in visual quality.

Aiming at the MTF evaluation limitations in visible imaging system, this paper introduces the minimum resolvable contrast. On the basis of MRC theoretical, the paper obtains subjective and objective methods of MRC measurement. Then MRC measurement based on neural network is put forward, which does not depend on subjective judgment of person. BP neural network is established and trained. Therefore, the network can replace human eyes to judge test patterns with different spatial frequencies and contrasts. Sony camera with 500 megapixels is selected in the experiment. Results show that MRC values of the objective measurement at all frequencies are less than those of the subjective measurement. The MRC Measurement has good stability.

Three-dimensional (3D) reconstruction based on a coded structured light method is a widely used technique to recover objects surfaces. This technique is based on projecting a pattern onto object to reconstruct and the illuminated scene is imaged by calibrated camera. The pattern creates an artificial features on the surfaces of objects and can be a correspondence points. The 3D information is reconstructed by triangulating the correspondence between projected pattern and decoded pattern in imaged scene. In this paper, 3D reconstruction method based on a coded structured light is proposed. The pattern is encoded in binary pattern and it has uniqueness in the search range to find correspondence points. The pattern is projected using an infrared laser and a diffraction optical element (DOE) for the invisible areas. Then, the illuminated scene is imaged by a sensitive camera. As a result, the disparity of correspondences between matched points draw a 3D surface information.

In order to improve the accuracy and reliability of temperature measurements, put forward a kind of high precision non-contact temperature measuring system, which is composed of the quartz temperature sensor array and microprocessors. The sensor used by the system is the quartz tuning fork non-contact temperature sensor based on polymer line, and the polymer line that is highly sensitive to temperature is as the sensitive element, in combination with quartz tuning fork resonator that is not sensitive to the temperature , and they form the high precision temperature sensor together. Compared with the traditional thermometry, set up a new temperature sensor array to measure temperature, which uses the Bayes estimation to fuse the data measured by the temperature sensor array, and calculate the accurate temperature measurements. The experimental results shown that the measurement precision of sensing system is 0.02℃, and the sensitivity is 71.5×10-6/℃within temperature range from 0°C to 100°C.The system has advantages of the high precision, and the good stability, and reliable performance, etc.

According to electric power harmonic measuring method of the digital sampling, the simulating was carried out in this paper, thus a kind of based on Newton iterative blind source analysis of Fast - three times the ICA algorithm was Put forward, and its convergence was proved, the simulation analysis of the real-time error and measurement algorithm was finished, its instantaneity and measuring of the algorithm was proved.

Space vector pulse wide modulation(SVPWM) have less harmonic than SPWM, it is easy to digitalize, and it is used in high voltage field more and more. In this paper, the principle is analyzed detailed, and simulated by Simulink, the simulation result show that the space vector pulse wide modulation scheme is feasible. The SVPWM power system hardware and software is designed at last.

Phase data modulating the intensity of fringe patters can provide useful information about object surface or shape using an optical fringe projection system. Measuring the phase map of recorded fringe pattern via an appropriate phase retrieval algorithm can lead as to the object shape information. Several physical and numerical parameters can affect the estimation of phase maps. We systematically studied effects of theses parameters on different phase analyses using a recently proposed directional wavelet algorithm (DWA) that is based on 2D continuous wavelet transform (2D-CWT), an advanced multiscale decomposition algorithm. We considered two types of parameters that are often presented in recorded intensity images: local spatial frequencies of data and modulation rate of fringe patterns. In this paper, we propose a study of the phase retrieval problem by comparing two different methods on numerical and experimental data, to improve the performance of the analysis process and then the accuracy of results.

The depth of working fluid level of an oil well is one of the critical parameters that reveal the condition of down-hole environment. It should be measured periodically and accurately to stabilize the usual operation of pumping units. Acoustic method is currently the most common and most effective measurement method. This paper discuss a kind of working fluid level measuring equipment, which is based on NI single-board RIO embedded real-time processing system. By combing existing analysis method with wavelet transform theory, this device can automatically calculate the sound speed in wellbore annulus, and recognize the waveform reflected by fluid level. Results show that the device can effectively detect the depth of the working fluid level, meeting the needs of normal production activities in oil field.

Fractional order PID controller has two degrees of freedom more than the integer order PID controller,and it has better control performance. The paper briefly describes the fractional order PID controller, and through Ziegler-Nichols type tuning rules, Padula & Visioli tuning rules and tuning by minimization methods to achieve fractional order PID controllers respectively. Additionally, the Oustaloup approximation methods are both introduced. Based on these methods, we have considered the steam pressure model and the vehicle wire control system. The simulation result shows that control system stability can be obtained by using the tuning-rules proposed in this paper.

For natural gas pipeline, it has a leak or not is critical. The most commonly problems in the pipeline leak detection methods are the difficulties to identify, inaccuracy to locate, thus, the natural gas pipeline detection is difficult to be applied, therefore, the use of neural network multi-sensor data fusion of the natural gas pipeline leak detection is particularly important. In this paper, the method is proposed based on RBF neural network and the data fusion of D-S evidence theory for detecting the pipeline leak. Extracting neural network's input parameters through wavelet denoising, then substitute them into neural network and calculate them by multi-sensor data fusion algorithm so as to acquire leaking information.

Iris biometric is most mature and secure authentication system as compared to other systems. Other authentication system does exist apart from biometrics such as PIN, password etc., which are not secure and more vulnerable to attacks and can be hacked or spoofed easily. Using, Iris biometric we can enhance overall performance in terms of accuracy, which is possible if and only if the pupil part must be removed perfectly and efficiently. Pupil part is unwanted part for our system. Pupil part is surrounded by iris part when extracted successfully we get two approximate concentric circle. Proposed scanning algorithm successfully extracts pupil part and defines iris part with less complexity and more efficiently. Iris part is consisting of patterns which are desired for authenticating a person and each patterns are represented in terms of feature vectors and stored in database. Proposed system focuses on feature extraction using five level decomposition technique implemented with haar, db2 and db4 and achieves high accuracy with reduced error rates. Due to reduced errors and considering lower half of iris part, proposed algorithm can be used for larger database such as for Aadhar because it takes less time for feature extraction and has less complexity with reduced mathematical burden on the system and improves good accuracy.

To realize rural road detection in complication environment, a method based on improved FCM (Fuzzy C-Means) and Hough transformation has been proposed. Firstly, the road image is transformed from RGB space into HIS space. Second, the initial clustering center is determined using climbing method looking for image saturation histogram peak in HIS color space and the road area is detected using improved FCM. Finally, after morphological image processing, road boundary is detected using Hough transformation. Experimental results show that the method presented in this paper has good ability to detect the road boundary. The method is robust and not influenced by shadow, water vestige, rainy days and other environmental factors.

This paper studies the multiple antenna system is a kind of cognitive radio can achieve full interference alignment method of multiplexing gain, sender in multiple users and multiple users topology results at the receiving end, the user system interference to primary user system to produce at the same time, also cause interference between each other, which reduces the primary user system and the size of user system, also can affect the user's system can't normal communication; This method is based on the interference to primary user system under the condition of limited power, maximize user system capacity; Through the analysis can see, interference alignment method through alignment interference subspace, can spare the relative number of signal subspace, the subspace can be used to transport the user system information and useful information of user's system, so as to eliminate the interference effect; The simulation results show that this method is compared with the previous beam forming algorithm, and can get more transfer rate, improve the capacity of second user's system.

Owing to the fact that fabric images exhibit strong periodic and their features are difficult to extract, an image mosaic method based on cloud model cellular automata is presented. The edge feature points from two images are extracted by cloud model cellular automata, and then the corresponding feature point pairs are got by the cross-correlation of the gray scale around the feature points. At last the images can be stitched by matched feature point pairs. The experiments show that this method can achieve image mosaic effectively.

Speaker recognition and speech recognition are both related. As against determining what was said, speaker recognition enables the automatic recognition of who is speaking based on the speaker’s voice’s unique characteristics. This paper presents a simple approach to text dependent speaker identification and is based on the Symlet wavelets for feature extraction. The extracted features are then classified using data mining algorithms. In this study, J48, Naïve Bayes and SVM are used for classifying the features.

This paper presents an architecture of Multiband Impulse Radio UWB (MB-IR-UWB) system based on Prolate Spheroidal Wave Function (PSWF) by binary time hopping pulse position modulation (TH-PPM) over multipath fading channel. In order to combat multipath fading, Rake receiver using maximal ratio combining (MRC) is applied to this system, which can successfully capture the unique multipath-rich property and multipath-clustering phenomenon of UWB channels. Through simulation, a comparison of system performance between different types of Rake receiver is analyzed. What’s more, considered multipath delay spread, a data rate estimation of the communicating system is given. Finally we get the conclusion that increasing data rate is best-achieved by increasing the number of subband instead of decreasing the symbol repetition period.

In this paper, we propose a novel on-line blind demodulation algorithm for Quadrature Amplitude Modulation (QAM). The proposed approach can demodulate QAM signal without constellation knowledge, and reach the theoretical optimal performance under the constraint of sequential detection. Numerical simulations are provided to validate the analytical results. Results show that when iterations are large enough, the bit error rate (BER) of the proposed approach can be close to the optimal demodulation algorithm.

Terahertz(THz) imaging provided a good contrast between skin cancer (basal cell carcinoma BCC) and healthy tissue in vitro and ex vivo owing to the high water content and strong absorption of cancer tissue at THz frequencies .Modeling the propagation of a THz pulse through BCC would contribute to revealing the diagnostic and potential therapeutic application value of THz radiation. In this letter, the healthy skin and BCC were modeled as Double Debye dispersive media and the model was incorporated into the FDTD method to simulate the propagation of a THz pulse. Furthermore, absorption properties were revealed by calculating average SAR value of the horizontal distribution and vertical distribution, which would help investigate the response below the skin surface.

The lack of information content in the images captured under bad weather dominates the reason of poor visibility more, than due to the overall contrast or other. This paper presents a novel method for enhancement of weather degraded images. The proposed enhancement method is based upon the logarithmic image processing (LIP) framework and human visual system (HVS). The proposed method uses multiple images of the same scene in order to increase the information content and produce a visibly enhanced resultant image. A region based contrast measure called the logarithmic region contrast measure (LRCM) for finding the optimal value of location based parameters of the proposed method and its comparison with other traditional contrast measures has been presented here in this paper. The experimental results on two data sets of weather degraded images are included here in this correspondence.

Aiming at medical image information security problem, this paper proposes a new robust zero-watermarking algorithm for medical volume data based on three-dimensional discrete wavelet transform, three-dimensional Fourier transform and hermite chaotic neural network . Firstly, it employs a novel hermite chaotic neural network to generate the pseudo-random chaotic sequence for scrambling. Secondly, three dimensional medical image is transformed by three-dimensional discrete wavelet transform and three-dimensional discrete Fourier transform. Then, select the transformed low and intermediate frequency coefficients symbol as medical volume data characteristics to structure zero-watermarking. The algorithm integrates hermite chaotic neural network and zero-watermarking technology, which is not confined to artificially selected region of interest. The watermarking extraction does not need the original image. And its security depends on the chaotic sequence complexity and unpredictability, solving the watermark embedding, extraction safety and efficiency. The simulation results show that the algorithm is simple to implement, with good robustness, security and invisibility.

To further improve the performance of artificial bee colony algorithm (ABC), an improved ABC (IABC) algorithm is proposed for global optimization via employing orthogonal initialization method. Furthermore, to balance the exploration and exploitation abilities, a new search mechanism is also designed. The performance of this algorithm is verified by using 27 benchmark functions. And the comparison analyses are given between the proposed algorithm and other nature-inspired algorithms. Numerical results demonstrate that the proposed algorithm outperforms the original ABC algorithm and other algorithms for global optimization problems.

Although using the singer's name to search for songs is the common function of music retrieval systems, people often only remember famous singers. For new singers, people often only remember that the singer’s voice is similar to a famous singer, but cannot recall his/her name. Among current studies, there is little discussion on the music search method by voice. Therefore, this study proposed a method of music search by similar vocal sound. Users can enter the name of a singer, and then the system can find the names of other singers with similar features in vocal sound. In the proposed method, the sound signal is processed by Discrete Fourier Transform, and the statistical value of frequency spectrum and formant is calculated as the characteristic value of the singer's vocal sound. Then, the similarity of the singer's vocal sound is calculated by the weighted squared Euclidean distance function to obtain similar singers' names, so as to search for music by using similar vocal sound. The experiment verified that the feature extraction method can realize the vocal sound-based search effectively.

Panoramic image registration is the core content of virtual three-dimensional system based on images. According to the shortage of fisheye images registration using the traditional image registration method, this paper put forward strategy of panoramic image registration, which was divided into three steps: project the fisheye images on the sphere, extract and match the feature points, transform the coordinates of panoramic images. The focus was improving the SIFT algorithm and RANSAC algorithm for image feature exaction and matching, whose target was to meet the demand of the fish-eye image registration. Finally, experiments were designed to verify the effectiveness of supposed algorithm and fisheye image registration, and the future direction of panoramic images registration was also pointed out.

Traditional smoke recognition based on HSI model or frame differentiation method can not reach the effect of real-time processing due to large amount of computation and requirement of too many images. A kind of automatic method for recognizing forest fire smoke based on gray bit plane aiming at to improve the real-time performance is proposed, analyze the gray information of forest fire smoke and use method combining gray bit plane with adaptive area of forest fire smoke, first, build gray bit plane of smoke image, and then calculate combining with the characteristics reflected by smoke information in each bit plane in order to improve the real-time performance and finally obtain the area of forest fire smoke. Experimental results show that the accuracy of recognizing area by analyzing and calculating gray bit plane is consistent with the traditional method, which takes a shorter time in processing and has actual application prospect to a certain degree.

This paper describes Support Vector Machine (SVM) and Artificial Neural Network (ANN) based recognition and classification of visual symptoms affected by fungal disease. Color images of fungal disease symptoms affected on cereals like wheat, maize and jowar are used in this work. Different types of symptoms affected by fungal disease namely leaf blight, leaf spot, powdery mildew, leaf rust, smut are considered for the study. The developed algorithms are used to preprocess, segment, extract features from disease affected regions. The affected regions are segmented using k-means segmentation technique. Color texture features are extracted from affected regions and then used as inputs to SVM and ANN classifiers. The texture analysis is done using Color Co-occurrence Matrix. Tests are performed to classify image samples. Classification accuracies between 68.5% and 87% are obtained using ANN classifier. The average classification accuracies have increased to 77.5% and 91.16% using SVM classifier.

Based on the unique geometric features of the ruled surface, a new efficient algorithm for ruled surface/plane intersection is proposed. The ruled surface is firstly dispersed into a set of line segments, and the ruled surface/plane intersection is transformed to the intersection of a group line segments with a plane. Then a set of ordered intersection points can be obtained by the proposed line/plane intersection algorithm. According to the serial number of every intersection point, all the intersection points are grouped and reorganized. Each group point corresponds to an intersection curve. All the intersection curves can be reconstructed by curve interpolation. Compared with the traditional tracing method, the proposed algorithm can avoid complex calculation including initial points searching and intersection points sorting, which is more efficient and stable.