Earticle

Traditional 2D face recognition based on optical (intensity or color) images faces many challenges, such as illumination, expression, and pose variation. In fact, the human face generates not only 2D texture information but also 3D shape information. In this paper, the objective is to investigate what contributions depth and intensity information make to the solution of face recognition problem when expression and pose variations are taken into account, and a novel system is proposed for combining depth and intensity information in order to improve face recognition performance. In the proposed approach, local features based on Gabor wavelets are extracted from depth and intensity images, which are obtained from 3D data after fine alignment. Then a novel hierarchical selecting scheme embedded in symbolic principal component analysis (Symbolic PCA) and AdaBoost learning is proposed to select the most effective and most robust features and to construct a strong classifier. Experiments are performed on the three datasets, namely,Texas 3D face database, Bhosphorus 3D face database and CASIA 3D face database, which contain face images with complex variations, including expressions, poses and long time lapses between two scans. The experimental results demonstrate the enhanced effectiveness in the performance of the proposed method. Since most of the design processes are performed automatically, the proposed approach leads to a potential prototype design of an automatic face recognition system based on the combination of the depth and intensity information in face images.

Thresholding based on variance analysis of gray levels histogram is a very effective technology for image segmentation. However, its performance is limited in conventional forms. In this paper, a novel method based on two-dimensional extension of within-class variance is proposed to improve segmentation performance. The two-dimensional histogram of the original and local average image is projected to one-dimensional space firstly, and then the minimum within-class variance criterion is constructed for threshold selection. The effectiveness of the proposed method is demonstrated by using examples from the synthetic and real-word images.

Script Identification is one of the challenging step in the Optical Character Recognition system for multi-script documents. In Indian and Non-Indian context some results have been reported, but research in this field is still emerging. This paper presents a research work in the identification of Gurmukhi and English scripts at word level. It also identifies English Numerals from Gurmukhi text. Gabor feature extraction is one of most popular method for script recognition. This paper presents a zone based gabor feature extraction technique. The given word image after normalization is divided into different zones of different sizes and then features from each of these zones are extracted in various directions using gabor filters. Script is then determined by using SVM classifier. The experimental tests carried out in the field of Gurmukhi and English Script recognition show that the proposed technique leads to improvement over the traditional Gabor feature extraction without zoning. In future, this can also be extended for other scripts.

According to grayscale changing characteristics of the brain CT, this paper improves region growth algorithm on the basis of in-depth analysis of brain CT image. In order to get the best combination of the features, the algorithm that extracts the shape feature considering the adjacent parts is proposed based on the idea of chart transferring tree in the tree structure comprehensively. Compared with the traditional extracting algorithms which aim at the whole picture or a single part, the experiments show that the proposed algorithm can obtain a higher diagnostic accuracy.

As a new biometric identification technology, the theory and application research of ear recognition has attracted more and more attention of scholars in recent years. Image-preprocessing and normalization of ear image is very important for the feature extraction. In this paper we apply the improved morphological filtering method to preprocessing the ear image. And we propose the angle normalization method by geometrical parameters. This method has the advantages of scaling invariance, translation invariance and rotation invariance. The normalization results are reasonable and good for later feature extraction.

The automobile service experts often assess the health condition of the motorcycles based on the sound produced by taking test rides. To be effective, this process of fault diagnosis needs to be automated. The purpose of this paper is to present a method for fault detection of motorcycles that employs the slopes of the pseudospectral segments as features. Further, the estimated pseudospectrum of a sound signal is divided into eight segments, and the slope of each segment is computed. Artificial neural network (ANN) classifier is used for classification. The experimental results show that the proposed method achieves satisfactory results with an average accuracy of 78% for healthy motorcycles and 89% for faulty motorcycles. The study can be extended to locate the faults in subsystems of vehicles. The proposed work finds applications in allied areas such as fault diagnosis of machinery, musical instruments, electronic gadgets etc.

Ultrasonic testing is an important flaw detection technology in nondestructive testing field. With its unique electronic scanning, dynamic deflection focusing (DDF), sectorial scanning characters and ultrasonic phased array (UPA) technology can be used for defecting detection of objects with curved face or complex structure. Thus, UPA technology has become a focus of ultrasonic testing research. In order to solve a kind of problem of the internal inspection, this paper analyze methods of phasing array imaging and make a simulation scheme. The simulation result shows that the fan imaging method proposed by this article is right, establishing basis for the real-time imaging test analysis of ultrasonic flaw detection.

The power frequency in 50Hz and its harmonic interference is one of the main noise sources during the acquisition process of surface electromyogram signals (SEMG), how to restrain interference power frequency and its harmonics effectively is one of the important issues on the EMG preprocessing. In this paper, we will determine the power frequency interference and its frequency of the harmonic through the signal acquisition, make the cross-correlation by the constructor and the collected data, and determine its magnitude. We can draw a conclusion that this method can effectively restrain power frequency interference of EMG during the acquisition process through the comparison of the power spectrum of de-noising before and after. The Experiments show that this method can effectively separate the power frequency interference during the acquisition process, and is suitable for the real-time processing of EMG.

This paper presents a new recognition algorithm for plant pathology images based on the Non-negative Matrix Factorization, the proposed algorithm is combined with optimal wavelet packet basis to recognize patterns and conduct data encoding in the internet of things oriented intelligent agricultural system. The experimental results show that the performance of the proposed recognition algorithm is far better than those of the principal component analysis and linear discriminant analysis, and the recognition rate are improved, on average, about 14.65% and 11.18% higher than the rates of the above algorithms respectively. The presented algorithm is characterized by the fast speed, high calculation accuracy and easy hardware implementation.

In present paper the effect of noise on fractal dimension of digital images has been tested. Since fractal dimension is a measure of texture which is property of neighbourhood, it is interesting to check how noise affects the values of fractal dimension. For this purpose, three standard digital images have been used and Gaussian noise, salt and pepper noise and speckle noise have been applied to these images to generate noisy images. The fractal dimension values of actual images and noisy images have been estimated and compared. Various aspects related to the estimation of fractal dimension of digital images are discussed for noisy and non-noisy images. Since variation of noise makes sense for a local window in which fractal dimension is estimated, it is required to observe the noise effect in the window.

New breakthroughs were taken from the research of affective curve extracting from sequence-concentrated functional Magnetic Resonance Imaging (fMRI) images, and the emotional responses of human brain were hidden in these fMRI dataset; the purpose of this paper is to acquire critical features from fMRI images. The fMRI experiments were given by a certain theme emotion stimuli; firstly, component operations under bilateral filtering were applied for fMRI images’ morphological segmenting which reduced the computational space, for that the calculation was not based on the whole brain space. Operated by Fast Fourier Transform (FFT), fMRI images relative to functional area of human brain were pre-processed. Finally, time series based Power Spectrum Density (PSD) was founded by using an improved shape preserving fitting algorithm, and affective curves were acquired subsequently. The results showed the effectiveness of the proposed methodologies in this paper by comparing with cubic fitting and 5-th polynomial fitting operations. Experimental results also showed that this method was effective and efficient; the shaper preserving model had the lowest residual error that reflected the brain's emotional response curve adequately. The proposed methods have potential applications in the study of human-machine emotion interactions.

3D-2D vehicle registration provides a new way for vehicle recognition, localization and tracking in traffic surveillance systems. In this paper we present two novel fitness functions to measure 3D-2D vehicle matching, where 3D wire-frame model is used. Unlike the existing vehicle registration methods, we group the model's wireframes into the important and unimportant ones in view of the disaccord between real vehicle and the wire-frame model. The important wireframes generally well fit the corresponding image edges, whereas the image edges corresponding to the unimportant wireframes may not exist due to the streamlined design of real vehicle. For more accurate matching, the fitting of the important wireframes is underlined in both two fitness functions. In the first fitness function, the larger weight coefficient is assigned to the fitting of the important wireframes; in the second fitness function, two different functions are used for the fitting of the model's wireframes instead of two different weight coefficients. Experiments on real traffic videos verify the correctness and robustness of the proposed fitness functions.

Medical image visualization system, which is of great value in medical research and clinic diagnosis, has been a focal field in recent years. IDL (Interactive Data Language) has a vast library of built-in math, statistics, image analysis and information processing routines, therefore, it has become an ideal software for interactive analysis and visualization of two or three dimensional scientific datasets. The principles and development are proposed to design a novel three dimensional visualization toolbox for medical images based on IDL platform using object oriented programming and object graphics techniques, with experimental results exhibited. It is demonstrated that our developed toolbox can carry out interactive volume and slice operations in both two and three dimensions effectively and efficiently, meanwhile, it has advantages of extensive applicability, friendly interaction, convenient extension and favorable transplantation, which is adequate for medical images processing and analysis.

Color image processing is widely used in Intelligent Transport System, but seldom used in recognition of roads and slopes collapse. The application can reduce time and efforts. And the roads and slopes segmentation is the first and key step of the recognition system, which is a challenging and difficult problem. One of the problems is the presence of different types of roads and slopes. In this paper, we propose a novel framework for segmenting road images in a hierarchical manner that can separate the following objects: road and slopes with or without collapse, sky, road signs, cars, buildings and vegetation from the images. Then the Region of Interests (ROIs), i.e. the roads and slopes, are obtained with the geometrical, location of the objects and statistical color features which are extracted based on L*a*b color space and Gabor filter. According to combination K-means clustering with region merging, connected-component algorithm and morphological operation, the roads and slopes are segmented. The hierarchical approach does not assume the roads are present in the same type and assume the road images can be captured from arbitrary angles. The experiments show that the approach in this paper can achieve a satisfied result on various road images.

This work presents a new method to virtually recover ancient Chinese paintings in electronic form. Two factors result in the degradation of ancient Chinese paintings: pigment fading and paper aging. Thus, the proposed method first uses guided filter to enhance the original painting image. Then, the semi-transparent stroke as the image foreground can be extracted from the background of painting image by adopting a novel bi-extracting method. Two independent maps that respectively produced by the saliency detection and the pseudo-alpha matting are fused to represent the foreground stroke. Finally, the ancient painting image is recovered by altering the color of the background. Experiments on a variety of ancient Chinese paintings show the robustness and accuracy of the proposed method.

Aiming to solve the problem of long table look-up time in table look-up of CAVLC (Context-based Adaptive Variable Length Coding) for H.264/ AVC, a quick table look-up algorithm based on hash query (TLHQ) is proposed to reduce table look-up time for CAVLC decoding in this paper. The basic idea of the new algorithm is that it uses query technology of hash table to rapidly determine code length based on the number of zero in code prefix, As a result, it can save a lot of time of table looking-up and code judging. Experimental results show that our supposed new algorithm could reduce about 20% table look-up time and save 1056 byte storage space for CAVLC decoding in H.264/AVC.

Aiming at partitioning an image into homogeneous and meaningful regions, automatic image segmentation is a fundamental but challenging problem in computer vision. It is well known that Fuzzy c-means (FCM) algorithm is one of the most popular methods for image segmentation. However, the FCM-based image segmentation algorithm must be manually estimated to determine cluster number by users. In this paper, we propose a novel cluster number adaptive fuzzy c-means image segmentation algorithm (CNAFCM) for automatically grouping the pixels of an image into different homogeneous regions when the cluster number is not known beforehand. We utilize the Grey Level Co-occurrence Matrix (GLCM) feature extracted at the image block level instead of at the pixel level to estimate the cluster number, which is used as initialization parameter of the following FCM clustering to endow the novel segmentation algorithm adaptively. We cluster image pixels according to their corresponding Gabor feature vectors to improve the compactness of the clusters and form final homogeneous regions. Experimental results show that proposed CNAFCM algorithm not only can spontaneously estimate the appropriate number of clusters but also can get better segmentation quality, in compare with those FCM-based segmentation methods recently proposed in the literature.

Robust degraded face recognition means the recognizer is robust to low resolution and blurry images and as well as other variations such as illumination, expression and et al. Such task is frequently encountered yet a challenging problem. In this paper, we propose appealing solutions to the task without any image reconstruction and without any blur type limitation. Short-Term Fourier Transform (STFT) is first conducted on face image and then two components relying on STFT are proposed: one is related to window size of STFT named scale and the other is face representation construction from STFT. The goal of the first component is to be robust to low resolution and blur. We propose a multi-scale competition strategy that extracts multiple descriptors corresponding to multiple window sizes of STFT and take the identity corresponding to maximum first candidate confidence as the final recognition result. The goal of the second component is to be robust to other variations. We explore the increased discrimination brought by joint coding and using of multiple frequencies. In particular, we propose a novel local descriptor in which infor-mation in local areas coming from two frequencies is jointly encoded and further multiple two-frequency-combinations are jointly utilized so as to construct a more discriminative and descriptive face representation. The experiments conducted on AR and Extended Yale B databases demonstrate that state-of-the-art performance has been achieved by mul-ti-scale competition strategy and the proposed novel face representation.

The existing structured light measurement technologies mainly focus on the single color objects, especially for measuring white object. Mainly because of in the process of three-dimensional measurement for color objects, color object’s surface has a great influence on color components of structured light, leading to the color of structured light changing, this will cause serious errors in the decoding process. To solve this problem, combined with the actual measurements for colored objects, this paper adopts a color Gray code for encoding and decoding structured light, and presents an obtaining technical for color components of structured light, which first through regression analysis builds a mathematical model, and then uses the least squares method for solving it, at last restores the color of the projected stripes to ensure the correctness of decoding, to achieve the measurement for color object and to improve the measurement accuracy. The experimental results show that this method has a good effect on decoding.

During the image edge detection with mathematical morphology, using single form and single scale of structural element will lose the information of other forms of elements. And the ability of anti-noise is weak. An adaptive algorithm for image edge detection based on multi-structural and multi-scale is proposed in this paper. For single form of structural element, we add weight coefficients on each result of different scale elements according to the inverse ratio of information entropy. For single scale of structural element, we add weight coefficients on each result of different form elements according to the direct ratio of information entropy. At last, we construct a series-parallel form of edge detector according to the algorithm process and get the final edge image after the fusion of edge images. The experiment shows that the improved algorithm not only can retain rich edge information but also has good performance to remove the noise.

Considering the traditional image match optimization algorithm is low efficiency, and ghosting artifact in the field of image stitching is a common problem and the elimination of it is not an easy task. In this paper, we presents that through cluster analysis for different scales of feature points to obtain the global and better stable affine transformation, and the matched feature points are filtered consistency by affine transformation, so it can eliminate the false matching points. An improved dynamic programming (DP) method is presented to find the best stitching line. The improved weighted average method is used to achieve smooth stitching results and eliminate intensity seam effectively. Experimental results show that the proposed algorithm has good result of removing false matching and ghosting artifacts, it is robust.

S-transform is a new time-frequency analysis method, which is deduced from short-time Fourier transform and continue Wavelet transform. It has much better performance than traditional time-frequency method. Therefore, in this paper, the basic principle of is briefly introduced and the relationships between is analyzed by theoretical derivation. According to the simulation experiments, the time-frequency space characteristics of short-time Fourier transform, Wigner-Ville distribution and S-transform are contrasted. As the results shown, the window of S-transform has a progressive frequency dependent resolution. So the S-transform has a great flexibility and utility in the processing of non-stationary signal. Compare with the time-frequency spectrum of three different analysis methods under various noise conditions, it is obvious that S-transform has much better anti-noise performance than that of traditional methods for non-stationary signal processing. Based on the superior time-frequency resolution, the S-transform spectrum can be used to describe the structure of incoming signal effectively.

Evidence theory is an important method for uncertainty reasoning. Because evidence theory can well process uncertain information, imprecise information, fuzzy information, and information without prior knowledge, it is wildly used in information fusion, expert system, fuzzy recognition, and intelligent decision system. This paper depend on original evidence theory fusion algorithm, put forward a new algorithm to process conflict information efficiently, this algorithm change the basic probability assignment of evidences by changing its reliability, and then get the ultimate decision making result by doing a new combination. The simulation results show the effectiveness of this algorithm, and it can give a better decision making results in processing conflict evidences.

Sigma-delta ACD has two main parts: analog modulator and digital filter, the performance of modulator determines the performance of sigma-delta ADC, so the design of modulator is very important. The paper introduces the principle of sigma-delta AD modulator with high accuracy and the applied over sampling technique, noise shaping technique and multi-bit quantizer technique. Determining the design scheme of modulator—three bits three orders CIFF(Cascade of integrators, feed forward form) structure, and it makes the behavior level verification for this scheme by Simulink tool in MATLAB. The simulation result shows that multi-bit quantizer modulator can get very high SNR, and based on this result it designs every part of the modulator circuit.

In this paper, a new simple and fast algorithm for image encryption is proposed. It exploits the scan patterns and function XOR in three standalone steps. The decryption procedure is similar to that of the encryption but in the reversed order. We implement and test the proposed algorithm using different sample image inputs and our experimental result and security analysis indicates the robustness and advantages of the new proposed algorithm. Using this algorithm it is possible to reproduce the original image with no loss of information for the encryption and decryption process. Algorithm is fast and simple enough to be comparable to other recent approaches, and it has passed all the security requirements and it is fast and secure to be used in very broad range of industrial applications.

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) in two directions of research. 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.

In this research, detection the meaning of pigpen chiper using a modification of the k-nearest neighbor algorithm. The detection process is by reading the image that contains the pigpen chiper, separating the pigpen chiper characters that exist in the image, and then detects the meaning of a pigpen chiper character one by one by comparing each pigpen chiper character with the images in the database, so it can be obtained an image to be used as a reference to detect the meaning of each pigpen chiper character. By using the proses, system can detect the meaning of pigpen chiper with 80% of accuracy. The calculation process is also very simple and fast.

The mean-removed vector quantization (MRVQ) scheme achieves good reconstructed image quality, but it requires a high number of bit rates. In this paper, we propose an improved MRVQ scheme. The compressed codes of MRVQ for an image block contain the block mean and the index recording the closest residual vector in the codebook. In the proposed scheme, the block mean values are encoded by the linear prediction technique followed by the Huffman coding technique. The MRVQ indices of the residual vectors are further compressed by the Huffman coding technique. From the experimental results, it is shown that a great deal of bit rate reduction is achieved by using the proposed scheme with acceptable image quality loss.

The aim of this work is to reduce the burden task on the DSP processor by transferring a parallel computation part on a configurable circuits FPGA, in automatic speech recognition module design, signal pre-processing, feature selection and optimization, models construction and finally classification phase are necessary. LMS filter algorithm that contains more parallelism and more MACs (multiply and Accumulate) operations is implemented on FPGA Virtex 5 by Xilings , MFCCs features extraction and DTW( dynamic time wrapping ) method is used as a classifier. Major contribution of this work are hybrid solution DSP and FPGA in real time speech recognition system design, the optimization of number of MAC-core within the FPGA this result is obtained by sharing MAC resources between two operation phases: computation of output filter and updating LMS filter coefficients. The paper also provides a hardware solution of the filter with detailed description of asynchronous interface of FPGA circuit and TMS320C6713-EMIF component. The results of simulation shows an improvement in time computation and by optimizing the implementation on the FPGA a gain in space consumption is obtained.

Histogram Equalization (HE) is a simple and effective image enhancement technique.But, it tends to change the mean brightness of the image to the middle level of the permitted range, and hence is not a very suitable for consumer product. While preserving the original brightness is essential to avoid annoying artefacts. To preserve brightness and to enhance contrast of images, numerous methods are introduced, but many of them present unwanted artefacts such as intensity saturation, over-enhancement and noise amplification. In the present paper, available histogram equalization based methods are reviewed and compared with image quality measurement (IQM)tools such as Absolute Mean Brightness Error (AMBE) to assess brightness preserving and Peak Signal-to-Noise Ratio (PSNR) to evaluate contrast enhancement.