Earticle

Hierarchical clustering (HC) algorithm can obtain good clustering results, but it needs large storage and computational complexity for large image processing. Anew color image segmentation algorithm based on mean shift and hierarchical clustering algorithm named MSHC is presented in this paper. MSHC algorithm preprocesses an input image by MS algorithm to form segmented regions that preserve the desirable discontinuity characteristics of image. The number of segmented regions, instead of the number of image pixels, is considered as the input data scale of HC algorithm. The proximity between each cluster is calculated to form the proximity matrix, and then ward algorithm is employed to obtain the final segmentation results. MSHC algorithm is employed on color image and medical image segmentation.

In this paper we describe our implementation of algorithms for face detection and recognition in color images under Matlab. For face detection, we trained a feedforward neural network to perform skin segmentation, followed by the eyes detection, face alignment, lips detection and face delimitation. The eyes were detected by analyzing the chrominance and the angle between neighboring pixels and, then, the results were used to perform face alignment. The lips were detected based on the analysis of the Red color component intensity in the lower face region. Finally, the faces were delimited using the eyes and lips positions. The face recognition involved a classifier that used the standard deviation of the difference between color matrices of the faces to identify the input face. The algorithms were run on Faces 1999 dataset. The proposed method achieved 96.9%, 89% and 94% correct detection rate of face, eyes and lips, respectively. The correctness rate of the face recognition algorithm was 70.7%.

Based on analyzing the related technologies of the original video retrieval, we proposed an algorithm of adaptive dual-threshold shot boundary detection for the game video of TV. At the beginning, we convert each frame to the HSV space, set high and low ratio of threshold according to the frame difference. Then, the average value of frame difference in the sliding window is calculated to receive the dual-threshold. Finally, the shot and the gradual are detected by comparing the frame difference and threshold.

Skeletonization is a crucial step in many digital image processing applications like medical imaging, pattern recognition, fingerprint classification etc. The skeleton expresses the structural connectivities of the main component of an object and is one pixel in width. Present paper covers the aspects of pixel deletion criteria in the skeletonization algorithms needed to preserve the connectivity, topology, sensitivity of the binary images. Performance of different skeletonization algorithms can be measured in terms of different parameters such as thinning rate, number of connected components, execution time etc. Present paper focuses on Peak Signal to Noise Ratio, number of connected components, execution time and Mean Square error on Zhang and Suen algorithm and Guo and Hall algorithm.

There are many problems in the current management of sports venues in China, such as venue service is too rigid, not the public service, leads to the resource waste of venues in your spare time; Venue fewer meets the requirement of training. Some managers' responsibility is consciousness, resulting in slow sports equipment is easily damaged and maintenance and so on. With the rapid development of our country economy, people's health consciousness is more and more strong, our sports venues have already can't adapt to the current situation. The rapid development of modern computer technology and network technology, have created favorable conditions for the network service platform of the development of sports venues, provides an important train of thought for the development of the sports venues of our country. This paper design the development of network service platform system of sports venues, the venue booking system, management system and financial system planning. Hope to improve the service level of venues, meet the exercise needs of the masses, and improve the economic benefit and social benefit.

Texture classification is important step in image processing and computer vision applications. The proposed method offers efficient way to classify the invariant texture using discrete shearlet transform and fuzzy logic. The texture features of an image are represented using shearlet energy features and shearlet co-occurrence features. These features are obtained from block based energy form of shearlet decomposed image using two levels of discrete shearlet transform with two directions and by varying the block size. Finally, the obtained parameters are used to classify the texture in an image using fuzzy logic classifier.

Previously a novel chaos M-ary modulation and demodulation method based on spatiotemporal chaos Hamilton oscillator was proposed and is applied in M-ary communication. Without chaos synchronization circumstance, the performance of communication has been improved in bandwidth efficiency, transmission efficiency and anti-white- noise performance compared with traditional communication method. Due to the working performance of zone partition demodulator is strictly connected with channel noise, it is necessary to study the influence of Hamilton phase trajectory deeply by the channel noise. In this paper, two methods for designing of the boundary of zone partition demodulator in additive white Gaussian noise are discussed, which are using statistics and curve fitting ways. Therefore, the problem about how to determine the boundary of zone partition demodulator in white noise is solved. Finally, the chaos M-ary communication system based on Hamilton oscillator is constructed and simulated in white noise channel. The result shows that the method proposed in this paper can improve the anti-white-noise performance of the whole communication system.

In this paper, we handle the problem of human action recognition by combining covariance matrices as local spatio-temporal (ST) descriptors and local ST features extracted densely from action video. Unlike traditional methods that separately utilizing gradient-based feature and optical flow-based feature, we use covariance matrix to fuse the two types of feature. Since covariance matrices are Symmetric Positive Definite (SPD) matrices, which form a special type of Riemannian manifold. To measure the distance of SPDs while avoid computing the geodesic distance between them, covariance features are transformed to log-Euclidean covariance matrices (LECM) by matrix logarithm operation. After encoding LECM by Locality-constrained Linear Coding method, in order to provide position information to ST-LECM features, spatial pyramid is used to partition the video frames, and the average-pooling-on-absolute-value function is implemented over each sub-frames. Finally, non-linear support vector machine is used as classifier. Experiments on public human action datasets show that the proposed method obtains great improvements in recognition accuracy, in comparison to several state-of-the-art methods.

Elimination of noise from the signal is the major task in signal processing applications. Wiener filter removes noise efficiently but it requires large number of computations and it was updated with speed issue with adaptive filter. Adaptive filter has several algorithms to remove noise from the signal. This paper performs cancellation of noise from the signal using wiener filter and adaptive filter algorithms namely LMS, NLMS and RLS algorithms in real time environment. All these methods are compared using several parameters like step size, mean and variance of noise, mean square error, signal to noise ratio, speed, no. of. Iterations etc. In the existence work, the authors have compared the performance of the wiener filter & LMS algorithm in real time environment with sinusoidal input. This paper is extended by comparing different adaptive filter algorithms with the input taken in real time environment. It is observed that RLS algorithm performs noise cancellation better than all other algorithms. But it has high degree of complexity & cost while NLMS algorithm has moderate speed of performance and it is quietly chosen for several applications.

Modulation type is one of the most important characteristics used in signal recognition. An algorithm to realize signal modulation identification is proposed in this paper. We applied wavelet transformation and STFT to the signal, and then used manifold learning method to reduce the high dimension and extracted the recognition feature. The proper threshold value was set as the classifier to achieve the purpose of recognizing 4 kinds of signals (MASK, MFSK, MPSK,QAM) in Gauss white noise environment. The algorithm requires priori signal information no other than signal-to-noise rate. Simulation result indicates the algorithm achieves good performance.

This paper mainly study traditional histogram image enhancement method. Aiming at the lack of image enhancement, a method based on human visual model was proposed, and the experimental results show that this algorithm has better performance than is a traditional histogram equalization method, which provided a good direction studying image enhancement algorithm.

Edge is the basic feature of image. Edges form the outline of an object. The need of edge detection is to find the discontinuities in depth, discontinuities in surface orientation, changes in material properties and variations in scene illumination. So edge detection is one of the most commonly used operations in image analysis and there are probably more algorithm for detecting edges. In this paper various edge detectors like Canny, Sobel, Roberts and Prewitt are compared. These operators are more susceptible to noise and do not give satisfactory result for face outline. For overcoming this disadvantage morphological method is studied and the result of edge detection using morphological method is compared with Canny edge detector, Sobel edge detector, Roberts edge detector and Prewitt edge detector.Wood and Glass Images are taken up as a special conditions for wider number of applications.

We introduce a fast and effective algorithm extreme learning machine (ELM) and apply it to image denoising. GA-ELM algorithm we proposed uses genetic algorithm(GA) to decide weights and bias in the ELM. It has better global optimal characteristics than traditional optimal ELM algorithm. In this paper, we used GA-ELM to do image denosing researching work. Firstly, this paper uses training samples to train GA-ELM as the noise detector. Then, we utilize the well-trained GA-ELM to recognize noise pixels in target image. And at last, an adaptive weighted average algorithm is used to recover noise pixels recognized by GA-ELM. Experiment data shows that this algorithm has better performance than other denosing algorithm.

Seam carving is a kind of content aware image retargeting algorithm and can be applied to resize and deliberately remove objects from digital images. Based on the observation that after applying an additional seam carving operation, the similarity, the energy relative error, and the difference of seam distance of original image are quite different from those of the seam-carved image, we propose and develop a new method for detecting seam carving or seam insertion of natural images without knowledge of the original image. First, we apply an additional seam carving operation to the testing image, then calculate similarity, energy relative error, and difference of seam distance between the testing image and its seam carved version. Last, we extract 11 dimensional features to detect seam carving operation to train a support vector machine classifier for recognizing whether an image is an original or it has been modified using seam-carving. Our experimental results demonstrate that our proposed forensic method achieves not only better detection rate but also lower dimensional features compared with other existing seam carved detection methods.

In this paper, the existing microscopic image segmentation method is studied, by comparing the segmentation result, the support vector machine (SVM) of microscopic image segmentation method has good precision, deserves further research.

Most researches on human behavior recognition are mainly based on the features of whole body motion. This paper proposed a hierarchical discriminative approach for recognizing human behavior based on limbs motion. The approach consists of feature extraction with mutual motion pattern analysis and discriminative behavior modeling in the hierarchical manifold space. A cascade CRF is introduced to estimate the motion patterns in the corresponding manifold subspace, and the trained SVM classifier is used to predict the behavior label for the current observation. The results on motion capure data prove the significance motion analysis of body parts, and the results on synthetic image sequences are also presented to demonstrate the robustness of the proposed algorithm.

The Project presents object tracking from videos based on template matching using Gaussian kernel with probability distribution function and mean shift algorithm. Initially, the target will be selected from chosen video sequence to track desired object in consecutive frames. The target will be utilized to determine the probability distribution function for similarity measurement between target and current processing frames. Here, Gaussian kernel function and its gradient are used here to find the PDF for corresponding templates. Similarity between two different images will be measured by weighted sum of Gaussian coefficients and PDFS. Mean shift approach used here to shifting the starting coordinates of template to find its similar features in consecutive frames to detect desired objects. The dissimilarity between the target model and target candidates will be expressed by a metric derived from Bhattacharyya coefficient. The project simulated results shows that moving object from video will be tracked accurately at different position and shape with help of templates in a considerable amount of time.

The whole world is rapidly integrated through the Internet, and the impact of the internet has penetrated into the whole society. Sports games method has strong entertainment, which can inspire the majority of students to participate in the initiative and enthusiasm, more conducive to the cultivation of the students' physical training. In this paper, the author analyzes the games teaching method of sports education by using network data ,the result shows that sports game method can significantly improve the quality of teaching, the student's quality of movement is better after using this teaching method. Combined with conventional basketball teaching, games teaching method can effectively relieve the tension, anger, fatigue, depression, energy and confusion of college students, improve the sense of self-esteem and significantly improve their mental health. So that, suitable teaching methods of physical education will inspire students' interest, enhance students' participation sports desire, and finally achieve the desired teaching effect.

EEG apparatus are expensive and bulky. Their real-time performance is weak, and EEG signals are easy to be distorted. In this paper, a low-cost portable EEG signal acquisition system based on DSP is developed. By a noninvasive method with bipolar leads, weak EEG signals are induced to the pre-processing circuits, where they will undergo multi-level amplifying and filtering. Then, the analog signals are converted into digital signals using ADS8320. These digitized signals are filtered in DSP (TMS320VC5509) so that the power interference and physiological artifacts are removed with LMS (Last Mean Square) algorithm and ICA (Independent Component Analysis) algorithm. Experimental results demonstrate that the system can acquire weak EEG signals in real time, display and save the processing results. The acquisition system has the advantages of usableness and portability, and helpful to the popularity of community-based and family-based EEG diagnostic equipments.

Image holistic scene understanding based on global contextual features and Bayesian topic model is proposed. The model integrates three basic subtasks: the scene classification, image annotation and semantic segmentation. The model takes full advantage of global feature information in two aspects. On the one side, the performance of image scene classification and image annotation are boosted by incorporating image global contextual features; On the other side, the performance of image semantic segmentation is also boosted by new superpixel region segmentation method and new superpixel regions and patch feature representation. 1) For image scene classification and image annotation: (1) We improve the feature engineering methods by using the PHOW proposed by Vedaldi [1]; (2) Furthermore, global contextual features are learned by semantic features. 2) For semantic segmentation: (1) We improve the super-pixel segmentation method by using UCM in the literature [2]; (2)We proposed new feature representation for super-pixel region and patches by incorporating DSIFT, texton filter banks, RGB color, HOG, LBP and location features. The experiments testify that model performance has raised on all three sub-tasks.

Satellite imageries which comprises of various multispectral spectral bands pertaining to spectral and spatial information of the images acquired by latest multispectral sensor technology are rapidly increasing day by day in the recent years for onboard satellite remote sensing applications. A lossy multispectral image compression is desired by the exploitation of the redundancies present in the spatial and spectral information while preserving the vital and crucial information of the image objects to a certain extent. In this paper a novel approach is proposed for lossy multispectral image compression which is an extension to the earlier existing algorithms. In this proposed method the multispectral images are first enhanced with interpolation based super resolution technique to estimate a hi-resolution (HR) image from a low-resolution (LR) input image. Secondly the decorrelated spectral bands transformed by discrete wavelet transform (DWT), which contain maximum entropy, are selected and these representative spectral bands are quantized and encoded using Improved SPIHT (ISPIHT) algorithm. The algorithm has been designed for the optimization of maximum coding efficiency and for high compression ratio of bits per pixel per band when compared with the well known compression techniques.

By sputtering and plasma spraying technology, a sapphire optical fiber blackbody cavity temperature sensor of tantalum (melting point is 2997℃) - zirconium oxide (melting point is 2715℃) thin film was developed. Static calibration system for the sensor to measure target temperature 1721℃ is also designed by using three oxygen (produced by water electrolysis) flame guns. When its impact resistant capacity is more than 50MPa, the corresponding temperature is 2802℃. A high power and high frequency modulation CO2 laser pulse is used as exciting source to heat it to 1500℃, and the dynamic calibration device is designed independently, thus this sensor in experimental results show a time response on the order of μs. Signal collection and transmission through fiber optic cable are proposed, which can meet the need of transient high temperature measurement in harsh environment.

For the distortion problem of traditional denoising threshold function exists, a seismic signal denoising method based on wavelet packet multi-threshold function was present. Seismic wave signals were done wavelet packet decomposition and wavelet packet decomposition coefficients were arranged in the order of magnitude of the frequency, the appropriate threshold criteria was selected on different frequency bands and threshold processing was done on different frequency bands. Synthetic seismic signals and real seismic data were done multi-threshold wavelet packet decomposition, which better remove high frequency random noise and retain useful signals. Experimental results showed that multi-threshold wavelet packet decomposition method can effectively remove noises and improve the resolution of seismic data, with better denoising effect.

Biological tissues have electrical conductivity and permittivity properties which depend on tissue composition, structure and health status. Bioelectrical properties can be utilized for non invasive disease diagnosis. Electrical Impedance Tomography (EIT) is a method for reconstructing the image of distribution of electrical conductivity and permittivity inside a volume from measurements made at the surface of the volume. EIT image reconstruction is an ill-posed problem that requires a priori information called regularization. The Total Variation (TV) regularization is often used in solving EIT inverse problem. In this paper, simulation has been carried out in noise free and noisy cases and TV regularized iterative Primal Dual Interior Point Method (PD-IPM) has been used to reconstruct the difference conductivity image.

Finite mixture model (FMM) with Gaussian distribution has been widely used in many image processing and pattern recognition tasks. This paper presents a new Student's-t mixture model (SMM) based on Markov random field (MRF) and weighted mean template. In this model, the Student's-t distribution is considered as an alternative to the Gaussian distribution due to the former is heavily tailed than Gaussian distribution, thus providing robustness to outliers. With the help of the weighted mean template, the spatial information between neighboring pixels of an image is considered during the learning step. In addition, the proposed method is able to impose the smoothness constraint on the pixel label by using MRF. Furthermore, an efficient energy function and a novel factor are applied in current model to decrease the computational complexity. Numerical experiments are presented on simulated and real world images, and the results are compared with other FMM-based models.

Animal image recognition, classification and retrieval from a database based on shape features are of research interest in image processing. This paper proposes a combined graph and texture based approach for recognition of animal contour images. The methodology uses texture features of the complete graph image obtained from the contour of an object. MPEG-7 database is used for testing. The recognition accuracy of 96% is achieved and is an improvement over the state-of-the-reported results. The work has potential applications in image retrieval from databases.

Weed identification is core of precision variable spray technology and weed information management system. Single type features are difficult to identify multi-class weeds in cotton fields. In this paper, multi-type feature fusion technique for weed identification is proposed. Firstly, multi-type features are extracted. In color feature extraction, FMS, SMS and TMS in HSI are extracted by color moment. In shape feature extraction, REC, RWL, CIR and SPH are extracted by geometric parameter method. In texture feature extraction, ASM, CON and COR are extracted by GLCM. Secondly, because feature dimension is too large, principle component analysis is used to reduce dimension to extract new features including COR, ASM, REC and two components. Finally, three comparative experiments including identification of five kinds of weeds, three kinds of weeds and two kinds of weeds are carried out. Experimental results show that method proposed in this paper is superior to state of the art and is suitable for identification of multi-class weeds. This method can also be applied in identifying weeds in other fields.