In order to improve chaos complexity and provide more reliable chaotic system for image encryption, and enhance the security of image encryption algorithm, a new image encryption algorithm based on two-dimensional inverse-trigonometric hyperchaotic system was proposed. Firstly, based on one-dimensional triangular function, a two-dimensional inverse-trigonometric hyperchaotic system was constructed. Compared with some two-dimensional chaotic systems, this system had wider chaotic range, more random iteration sequences and better ergodicity by simulation experiments about bifurcation diagram and Lyapunov exponent. Then based on the proposed chaotic system, the "scrambling-diffusion" strategy was designed and different keys were given, which were used to generate different hyperchaotic sequences. The image matrix was scrambled without repetition by hyperchaotic sequences, then the scrambled sequence were shifted and diffused. So the ciphertext was obtained by looping thrice. Finally, histogram analysis, key space analysis, correlation analysis of adjacent pixels, plaintext sensitivity analysis and information entropy analysis were carried out. The test values of Number of Pixels Change Rate (NPCR) and Unified Average Changing Intersity (UACI) of ciphertext images were very close to their ideal expected values. The test results of information entropy were about 7.997, which was also very close to the expected value of 8. The experimental results show that the image encryption system has more reliable security, stronger ability to resist attacks, and had a good application prospect in the field of image security.

Focusing on the issue of extracting title and career attributes from unstructured text for specific person, an distant supervision and pattern matching based method was proposed. Features of personal attributes were described from two aspects of string pattern and dependency pattern. Title and career attributes were extracted by two stages. At first, both distant supervision and human annotated knowledge were used to build high coverage pattern base to discover and extract a candidate attribute set. Then the literal connections among multiple attributes and dependency relations between the specific person and candidate attributes were used to design a filtering rule set. Test on CLP-2014 PAE share task shows that the F-score of the proposed method reaches 55.37%, which is significantly higher than the best result of the evaluation ( F-measure 34.38%), and it also outperforms the method based on supervised Conditional Random Field (CRF) sequence tagging method with F-measure of 43.79%. The experimental results show that by carrying out a filter process, the proposed method can mine and extract title and career attributes from unstructured document with a high coverage rate.

RAMCloud stores data using log segment structure. When large amount of small files store in RAMCloud, each small file occupies a whole segment, so it may leads to much fragments inside the segments and low memory utilization. In order to solve the small file problem, a strategy based on file classification was proposed to optimize the storage of small files. Firstly, small files were classified into three categories including structural related, logical related and independent files. Before uploading, merging algorithm and grouping algorithm were used to deal with these files respectively. The experiment demonstrates that compared with non-optimized RAMCloud, the proposed strategy can improve memory utilization.

Traditional image object detection algorithm based on local feature is sensitive to rotation and occlusion; meanwhile, it also obtains low detection precision and speed in many cases. In order to improve the performance of this algorithm, a new image objects detection method applying objects’ local feature to sparse representation theory was introduced. Employing supervised random tree method to learn local features of sample images, a dictionary could be formed. The combination of sub-image blocks of test image and well trained dictionary in first stage could predict the location of the object in the test image, in this way it could obtain a sparse representation of the test image as well as the object detection goal. The experimental results demonstrate that the proposed method achieves robust detection results in rotation, occlusion condition and intricate background. What’s more, the method obtains higher detection precision and speed.

To solve the low efficiency and precision of the traditional methods, a near-duplicate image detection algorithm based on Bag-of-words and Hash coding was proposed in this paper. Firstly, a 500-dimensional feature vector was used to represent an image by Bag-of-words; secondly, feature dimension was reduced by Principal Component Analysis (PCA) and Scale-Invariant Feature Transform (SIFT) and features were encoded by Hash coding; finally, dynamic distance metric was used to detect near-duplicate images. The experimental results show that the algorithm based on Bag-of-words and Hash coding is feasible in detecting near-duplicate images. This algorithm can achieve a good balance between precision and recall rate: the precision rate can reach 90%-95%, and entire recall rate can reach 70%-80%.

In order to solve the problems that scene matching cannot be applied to navigation and Inertial Navigation System (INS) pose information cannot be effectively used, a new approach based on the digital terrain elevation model was proposed to estimate the absolute position and pose for aerial vehicle. Firstly, the stereo pairs were captured by the airborne camera system and the information of flight speed was acquired by sensors; secondly, the terrain information under the vehicle was obtained by modifying the dual camera motion model; finally, a method used to match the on-board terrain elevation model data based on the rigid-body constraint of the 3D reconstruction results was proposed, which was used to estimate the position and pose of the vehicle in the world coordinate system. The simulation results demonstrate that the improved dual camera motion model achieves better precision, making the estimation of position and pose in the world coordinate system more effective and efficient.

A designed goal for the general testing system on the basis of classified the network equipment testing was put forword.Through analyzed this goal,several key problems to be settled were proposed,and then answers to each question was put forward.After combined these answers,the final designed model for the testing system was proposed.At last,test items that could be tested in this system were introduced.After the practical testing,a comprehensive assessment was given.

 A matching algorithm based on the negative selection for anomaly detection was presented in this paper. In the algorithm the effects of position between two temporal sequence to matching degree were considered. So it could distinguish accurately self and non-self, and reduced the size of detective set effectively. Using normal sequence calls, the initial detective set was created, and the detective set was extended by learning, according to the proportion of anomaly temporal sequence to judge whether this sequence was anomaly. Finally, the results of experiment was given.