In order to realize accurate management of yaks during the process of yak breeding, it is necessary to recognize the identities of the yaks. Yak face recognition is a feasible method of yak identification. However, the existing yak face recognition algorithms based on neural networks have the problems such as too many features in the yak face dataset and long training time of neural networks. Therefore, based on the method of transfer learning and combined with the Visual Geometry Group (VGG) network and Convolutional Neural Network (CNN), a Parallel CNN (Parallel-CNN) algorithm was proposed to identify the facial information of yaks. Firstly, the existing VGG16 network was used to perform transfer learning to the yak face image data and extract the yaks' facial information features for the first time. Then, the dimensional transformation was performed to the extracted features at different levels, and the processed features were inputted into the parallel-CNN for the secondary feature extraction. Finally, two separated fully connected layers were used to classify the yak face images. Experimental results showed that Parallel-CNN was able to recognize yak faces with different angles, illuminations and poses. On the test dataset with 90 000 yak face images of 300 yaks, the recognition accuracy of the proposed algorithm reached 91.2%. The proposed algorithm can accurately recognize the identities of the yaks, and can help the yak farm to realize the intelligent management of the yaks.

In order to optimize the Mel-Frequency Cepstral Coefficient (MFCC) feature extraction algorithm, improve the recognition accuracy of respiratory sound signals, and achieve the purpose of identifying Chronic Obstructive Pulmonary Disease (COPD), a feature extraction algorithm with the fusion of MFCC based on Hilbert-Huang Transform (HHT) and short-term Energy, named HHT-MFCC+Energy, was proposed. Firstly, the preprocessed respiratory sound signal was used to calculate the Hilbert marginal spectrum and marginal spectrum energy through HHT. Secondly, the spectral energy was passed through the Mel filter to obtain the eigenvector, and then the logarithm and discrete cosine transform of the eigenvector were performed to obtain the HHT-MFCC coefficients. Finally, the short-term energy of signal was fused with the HHT-MFCC eigenvector to form a new feature, and the signal was identified by Support Vector Machine (SVM). Three feature extraction algorithms including MFCC, HHT-MFCC and HHT-MFCC+Energy were combined with SVM to recognize the respiratory sound signal. Experimental results show that the proposed feature fusion algorithm has better respiratory sound recognition effect for both COPD patients and healthy people compared with the other two algorithms:the average recognition rate of the proposed algorithm can reach 97.8% on average when extracting 24-dimensional features and selecting 100 training samples, which is 6.9 percentage points and 1.4 percentage points higher than those of MFCC and HHT-MFCC respectively.

Aiming at the problem of low detection rate of depth feature based object detection method Faster R-CNN （Faster Region-based Convolutional Neural Network） in flame detection tasks, a color-guided anchoring strategy was proposed. In this strategy, a flame color model was designed to limit the generation of anchors, which means the flame color was used to limit the generation locations of the anchors, thereby reducing the number of initial anchors and improving the computational efficiency. To further improve the computational efficiency of the network, the masked convolution was used to replace the original convolution layer in the region proposal network. Experiments were conducted on BoWFire and Corsician datasets to verify the detection performance of the proposed method. The experimental results show that the proposed method improves detection speed by 10.1% compared to the original Faster R-CNN, has the F-measure of flame detection of 0.87 on BoWFire, and has the accuracy reached 99.33% on Corsician.The proposed method can improve the efficiency of flame detection and can accurately detect flames in images.

An online multiple target tracking method for the aerial infrared targets was proposed based on the hierarchical data association to solve the tracking difficulty caused by the high similarity, large number and large false detections of the targets in star background. Firstly, according to the characteristics of the infrared scene, the location features, gray features and scale features of the targets were extracted. Secondly, the above three features were combined to calculate the preliminary relationship between the targets and the trajectories in order to obtain the real targets. Thirdly, the obtained real targets were classified according to their scales. The large-scale target data association was calculated by adding three features of appearance, motion and scale. The small-scale target data association was calculated by multiplying the two features of appearance and motion. Finally, the target assignment and trajectory updating were performed to the two types of targets respectively according to the Hungarian algorithm. Experimental results in a variety of complex conditions show that:compared with the online tracking method only using motion features, the proposed method has the tracking accuracy improved by 12.6%; compared with the method using multi-feature fusion, the hierarchical data correlation of the proposed method not only improves the tracking speed, but also increases the tracking accuracy by 19.6%. In summary, this method not only has high tracking accuracy, but also has good real-time performance and anti-interference ability.

Generally, there are inevitable disturbances in time series data, such as inherent uncertainties and external interferences. To detect outlier in time series data with time-varying disturbances, an approach based on prediction model using Gaussian Processes was proposed. The monitoring data was decomposed into two components:the standard value and the deviation term. As the basis of model for the ideal standard value without any deviation, Gaussian processes were also employed to model the heteroscedastic deviations. The posterior distribution of predicted data which is analytically intractable after introducing deviation term was approximated by variational inference. The tolerance interval selected from posterior distribution was used for outlier detection. Verification experiments were conducted on the public time series datasets of network traffic from Yahoo. The calculated tolerance interval coincided with the actual range of reasonable deviation existing in labeled normal data at various time points. In the comparison experiments, the proposed model outperformed autoregressive integrated moving average model, one-class support vector machine and Density-Based Spatial Clustering of Application with Noise (DBSCAN) in terms of F1-score. The experimental results show that the proposed model can effectively describe the distribution of normal data at various time points, achieve a tradeoff between false alarm rate and recall, and avoid the performance problems caused by improper parameter settings.

Video transcoding is applied in the field of Internet video coding. When the original video is transcoded multiple times, only the distortion between the input video and the output video can be calculated and the distortion between the output video and the original video can not be learned. Here an algorithm for estimating the distortion between the output video and the original video was proposed to control the quality of the output program. Firstly, the superposition of distortion caused by multiple lossy transcoding was analyzed to derive the lower limit of total distortion. Then the probability method was exploited to make an estimation on the distortion between the original video and the final output video. Finally, the least square fitting was used to correct the estimation according to the prediction error. Experimental results demonstrate that the proposed algorithm can accurately estimate the distortion with the prediction error of 0.02dB, 0.05dB and 0.06dB for Y, U and V components on average respectively after correction.

Conventional approaches for Common Weights (CW) generation in Data Envelopment Analysis (DEA) are either non-linear or scale-relevant. To solve this problem, according to the demand of military training performance evaluation, a new method was proposed to generate CW in DEA. The new method took DEA efficient units as the basis of calculation. Firstly, training data were normalized, and then multi-objective programing was employed for CW generation, which can lead to a fairer and more reasonable ranking of performances. The proposed method is not only linear, but also scale-irrelevant. Lastly, a military application illustrates that the proposed method is scientific and effective.

Aiming at the problem that the classification accuracy in malware behavior analysis system was low,a malware classification method based on Support Vector Machine (SVM) was proposed. First, the risk behavior library which used software behavior results as characteristics was established manually. Then all of the software behaviors were captured and matched with the risk behavior library, and the matching results were converted to data suitable for SVM training through the conversion algorithm. In the selection of the SVM model, kernel function and parameters (C,g), a method combining the grid search and Genetic Algorithm (GA) was used to search optimization after theoretical analysis. A malware behavior assessment system based on SVM classification model was designed to verify the effectiveness of the proposed malware classification method. The experiments show that the false positive rate and false negative rate of the system were 5.52% and 3.04% respectively. It means that the proposed method outperforms K-Nearest Neighbor (KNN) and Naive Bayes (NB); its performance is at the same level with the BP neural network, however, it has a higer efficiency in training and classification.

Aiming at the lack of semantic description for Service-Level Agreement (SLA) elements used in negotiation and the negotiation process in the SLA auto-negotiation, a negotiation mechanism based on the semantic Web technology was proposed, At first, a negotiation ontology named Osn was proposed, which was used for the description of SLA elements directly used in negotiation;the mapping function and the evaluation function of negotiation for these SLA elements were designed and described in this Osn, and the formal description of the main concepts and the relationship between these concepts was given based on description logic to provide a satisfiable semantic model for the Osn. Then a bargain model was put forward for SLA negotiation, and it was illustrated that a Pareto optimal offer could be generated by adopting this model through the proof of the related proposition and theorem;the service ontology was designed for SLA negotiation based on the mapping between OWL-S and Unified Modeling Language (UML) using this bargain model. The result of case study shows that the knowledge can form the sequence of offers which satisfied the need to maximize the interest of negotiation participants. It is illustrated that Osn can provide the service ontology with the parameter type support for the negotiation of an arbitrary SLA;the SLA negotiation oriented bargain model can generate the SLA accepted by both negotiation participants.

For the users' privacy security in electronic transactions, an electronic transaction scheme was proposed to protect the users' privacy. The scheme combined the oblivious transfer and ElGamal signature, achieved both traders privacy security in electronic transactions. A user used a serial number to choose digital goods and paid the bank anonymously and correctly. After that, the bank sent a digital signature of the digital goods to the user, then the user interacted with the merchant obliviously through the digital signature that he had paid. The user got the key though the number of exponentiation encryption, the merchant could not distinguish the digital goods ordered. The serial number was concealed and restricted, so the user could not open the message with the unselected serial number, they could and only could get the digital goods they paid. Correctness proof and security analysis shows that the proposed scheme can protect both traders mutual information in electronic transactions and prevent merchant's malicious fraud. The scheme has short signature, small amount of calculation and dynamic changed keys, its security is strong.

In cognitive Orthogonal Frequency Division Multiplexing (OFDM) systems, to avoid interference to Primary Users (PU), the transmission power of Cognitive Users (CU) need to be controlled and allocated. Since the transmission power can not be allocated legitimately and the data transmission rate can not be improved effectively, a power allocation algorithm of double factor binary search optimization was proposed on the basis of traditional water-filling power allocation algorithm. In the presented algorithm, the interference temperature limit on the cognitive user channel was taken into account. Firstly, a surplus function was introduced under the total power constraints. Secondly, because of the monotonicity of the surplus function, the accurate values of Lagrangian multipliers could be attained through the double binary search iteration method. Finally, the power allocation of the sub-channels was conducted through the values of Lagrangian multipliers. The simulation results show that the proposed algorithm can effectively use the spectrum hole between primary users. The data transmission rate of the cognitive users can be maximized under both total power constraints and Interference Temperature (IT) constraints. The data transmission rate is approaching to the traditional water-filling algorithm. Compared with the total power average control algorithm and the interference temperature average control algorithm, the data transmission rate of the presented algorithm is obvious higher, which exceeds about 4×105b/s under the same circumstance. Moreover, the algorithm has less processing time and reflects a good robustness.

Node deployment is a basic problem in sensor networks, which directly relates to the performance of the entire network. Most existing researches on sensor network node deployment are for the case of twodimensional planar and three dimensions space, but very few researches for threedimensional surface deployment scenario. This paper proposed an algorithm of optimal surface deployment in wireless sensor networks. First by mathematical or differential geometry method for threedimensional surface it constructed mathematical model, and then through the centroid of the threedimensional surface Voronoi subdivision partitions, an error function was proposed to evaluate the superiority of deployment method. Finally compared with other surface deployment methods, the performance of the proposed algorithm in this paper is superior.

The architecture of software and hardware of the embedded CAN-Ethernet gateway were introduced, and the principle, the designing methods and technoloques of the CAN Device Driver in uClinux were described. According to the features of the CAN protocol, data package was classified into four groups with different real-time request; the structure of multi-frame was proposed to satisfy the request of sending mass data; the data structure and the method of management for the buffer of the CAN Device Driver were designed to improve the capability of communication.