As a fundamental and key task in computer-aided diagnosis, medical image segmentation aims to accurately recognize the target regions such as organs, tissues and lesions at pixel level. Different from natural images, medical images show high complexity in texture and have the boundaries difficult to judge caused by ambiguity, which is the fault of much noise due to the limitations of the imaging technology and equipment. Furthermore, annotating medical images highly depends on expertise and experience of the experts, thereby leading to limited available annotations in the training and potential annotation errors. For medical images suffer from ambiguous boundary, limited annotated data and large errors in the annotations, which makes it is a great challenge for the auxiliary diagnosis systems based on traditional image segmentation algorithms to meet the demands of clinical applications. Recently, with the wide application of Convolutional Neural Network (CNN) in computer vision and natural language processing, deep learning-based medical segmentation algorithms have achieved tremendous success. Firstly the latest research progresses of deep learning-based medical image segmentation were summarized, including the basic architecture, loss function, and optimization method of the medical image segmentation algorithms. Then, for the limitation of medical image annotated data, the mainstream semi-supervised researches on medical image segmentation were summed up and analyzed. Besides, the studies related to measuring uncertainty of the annotation errors were introduced. Finally, the characteristics summary and analysis as well as the potential future trends of medical image segmentation were listed.

In order to solve the problems of insufficient consideration of the traffic flow characteristics and the low accuracy of the prediction, a short-term prediction method of urban road traffic flow based on spatio-temporal node selection and deep learning was proposed. Firstly, the characteristics of traffic flow were analyzed in theory and data representation to obtain its spatial characteristics, and temporal characteristics and candidate spatio-temporal nodes set. Secondly, the set of candidate spatio-temporal nodes was determined according to the reachable range of traffic flow, and the fitness was calculated by taking the inverse of the sum of squares of errors as the objective function. In the historical training set, genetic algorithm and Back Propagation Neural Network (BPNN) were used to select spatio-temporal nodes, and the final spatio-temporal nodes and BPNN structure were obtained. Finally, the measured values of the selected spatio-temporal nodes were taken as the input of BPNN in the working set to obtain the predicted values. The experimental results show that compared with only using data of adjacent spatio-temporal nodes, using other time node ranges, Support Vector Machine (SVM) and Gradient Boosting Decision Tree (GBDT), the proposed model has a slight reduction in Mean Absolute Error （MAE） and Mean Absolute Percentage Error (MAPE), which are 10.631 6 and 14.275 8%, respectively; and 0.257 3和0.999 1 percentage points lower than those by using adjacent spatio-temporal nodes.

To improve performance of current Non-Reference Image Quality Assessment (NR-IQA) methods, a no-reference image quality assessment algorithm with enhanced adversarial learning was proposed under the latest deep Generative Adversarial Network (GAN) technology. In the proposed algorithm, the adversarial learning was strengthened by improving the loss function and the structure of the network model, so as to output more reliable simulated "reference images" to simulate human visual comparison process as the Full-Reference Image Quality Assessment (FR-IQA) method does. First, the distorted image and undistorted original image were input to train the network model based on the enhanced adversarial learning. Then, a simulated image of the image to be tested was output from the trained model, and the deep convolution features of the reference image were extracted. Finally, the deep convolution features of reference image and the distorted image to be tested were merged and input into the trained quality assessment regression network, and the assessment score of the image was output. Datasets LIVE, TID2008 and TID2013 were used to perform the experiments. Experimental results show that the overall subjective performance on image quality assessment of the proposed algorithm is superior to those of the existing mainstream algorithms and is consistent with the performance of the human subjective assessment.

Focusing on the issue that classic semi-supervised image segmentation methods have difficulty in accurately segmenting scattered or small regions, a semi-supervised segmentation algorithm based on label prior and Laplacian Coordinates (LC) was proposed. Firstly, the Laplacian coordinates model was extended, and further the relationship between unlabeled pixels and labeled pixels accurately characterized by introducing the label prior. Secondly, based on the derivation of matrix equation, the posterior probability that the pixel belongs to the label was able to be effectively estimated, thus achieving the segmentation of the image. Thanks to the introduction of the label prior, the algorithm was more robust to the segmentation of scattered and small regions. Lastly, the experimental results on several public semi-supervised segmentation datasets show that the segmentation accuracy of the proposed algorithm is significantly improved compared with that of the Laplacian coordinates algorithm, which verifies the effectiveness of the proposed algorithm.

Concerning the problem that traditional crack detection algorithm based on percolation model has low efficiency and detection results are prone to fracture, a crack detection algorithm based on multi-factor decision and percolation model was proposed. Firstly, an improved algorithm of accelerating crack inspection based on percolation model was proposed, which improves the efficiency of percolation processing by reducing a large number of redundant pixel points involved in percolation processing. Secondly, the extracted percolation points were used to percolation processing. Finally, a multi-factor decision connection algorithm based on crack orientation was proposed. In the algorithm, the rationality of crack connection was analyzed by four decision factors to improve the accuracy of crack connection. Different morphological crack images with different interfering objects in background were used in experiments. Compared with traditional percolation model detection algorithm and original algorithm of accelerating crack inspection based on percolation model and skeleton connection algorithm, the number of percolation points of the proposed algorithm was reduced by an average of 99.7% and 38.1%, respectively. The precision was increased by an average of 60.5% and 6.4%, respectively, and the recall was increased by an average of 10.5% and 4.0%, respectively. The experimental results show that the proposed algorithm can significantly improve the efficiency of percolation processing and improve the accuracy of crack detection.

In order to improve the robustness of Non-negative Matrix Factorization (NMF) algorithm for speech enhancement in different background noises, a speech enhancement algorithm based on Sparsity-regularized Robust NMF (SRNMF) was proposed, which takes into account the noise effect of data processing, and makes sparse constraints on the coefficient matrix to get better speech characteristics of the decomposed data. First, the prior dictionary of the amplitude spectrum of speech and noise were learned and the joint dictionary matrix of speech and noise were constructed. Then, the SRNMF algorithm was used to update the coefficient matrix of the amplitude spectrum with noise in the joint dictionary matrix. Finally, the original pure speech was reconstructed, and enhanced. The speech enhancement performance of the SRNMF algorithm in different environmental noise was analyzed through simulation experiments. Experimental results show that the proposed algorithm can effectively weaken the influence of noise changes on performance under non-stationary environments and low Signal-to-Noise Ratio (SNR) (<0 dB), it not only has about 1-1.5 magnitudes improvement in Source-to-Distortion Ratio (SDR) scores, but also is faster than other algorithms, which makes the NMF-based speech enhancement algorithm more practical.

Aiming at the communication problem of crossing the rack switches for a large amount of intermediate data generated after the Map phase in the MapReduce process, a new optimization method was proposed for the map-intensive jobs. Firstly, the features from the pre-running scheduling information were extracted and the data communication activity was quantified. Then naive Bayesian classification model was used to realize the classification prediction by using the historical jobs running data to train the classification model. Finally, the jobs with active intermediate data communication process were mapped into the same rack to keep communication locality. The experimental results show that the proposed communication optimization scheme has a good effect on shuffle-intensive jobs, and the calculation performance can be improved by 4%-5%. In the case of multi-user multi-jobs environment, the intermediate data can be reduced by 4.1%. The proposed method can effectively reduce the communication latency in large-scale data processing and improve the performance of heterogeneous clusters.

To improve the universality of image feature extraction, an image feature extraction method based on improved Nonnegative Matrix Factorization (NMF) was proposed. Firstly, considering the practical significance of extracted image features, NMF model was used to reduce the dimension of image feature vector. Secondly, in order to represent the image by a small number of coefficients, a sparse constraint was added to the NMF model as one of the regular terms. Then, to make the optimized feature have better inter-class differentiation, the clustering property constraint would be another regular term of the NMF model. Finally, through optimizing the model by using gradient descent method, the best feature basis vector and image feature vector could be acquired. The experimental results show that for three image databases, the acquired features extracted by the improved NMF model are more conducive to correct image classification or identification, and the False Accept Rate (FAR) and False Reject Rate (FRR) are reduced to 0.021 and 0.025 respectively.

Since the problem of trajectory privacy violation and homogeneous semantic location attack of moving objects in road network environment is very serious, a Privacy-aware Trajectory Data Collection (PTDC) algorithm was proposed. Firstly, through visits' entropy of Points Of Interests (POI), the sensitivity of each POI was computed; secondly, based on the mixture distance of sensitivity and Euclidean distance, θ-weight was defined and a weighted model of vertices and edges in the network environment was established to reach a k-θ-D anonymity, which can resist the semantic location homogeneity attack; finally, based on the bread-first traversal algorithm of undirected graph, an anonymous algorithm was proposed to satisfy the semantic difference of POIs, so that user's sensitive sampling location was replaced by an anonymous region. Data utility caused by PTDC algorithm was theoretically evaluated. A set of experiments were implemented to test PTDC algorithm, and compare it with the privacy-preserving algorithm named YCWA (You Can Walk Alone) in free space. In theory, the privacy level of YCWA algorithm was lower than PTDC algorithm. The experimental results show that the PTDC algorithm has an average information loss of about 15%, and average range count query error rate of about 12%, which performs slightly worse than YCWA algorithm, while the running time of PTDC algorithm is less than 5 seconds, which is much better than YCWA algorithm. PTDC algorithm meets the needs of real-time online data collection.

Focusing on the issues that the sparseness of the data obtained after Non-negative Matrix Factorization (NMF) is reduced and the computing scale increases rapidly with the increasing of training samples, an incremental learning algorithm based on graph regularized non-negative matrix factorization with sparseness constraints was proposed. It not only considered the geometric structure in the data representation, but also introduced sparseness constraints to coefficient matrix and combined them with incremental learning. Using the results of previous factorization involved in iterative computation with sparseness constraints and graph regularization, the cost of the computation was reduced and the sparseness of data after factorization was highly improved. Experiments on both ORL and PIE face recognition databases demonstrate the effectiveness of the proposed method.

Combined ID-based partially blind signature and server-aided verification signature, a partially blind signature scheme with ID-based server-aided verification was presented to overcome the shortcomings of ID-based partially blind signature schemes such as strong security assumption and high computation cost. Most computing tasks of signature verification were accomplished by a server, and it greatly reduced computational overhead of verifier. Based on bilinear mapping, a partially blind signature scheme with specific ID-based server-aided verification was proposed. This scheme was proven to be secure in the standard model. Analysis results show that the proposed scheme greatly reduces computational complexity of signature verification. The proposed scheme is more efficient than Li's scheme (LI F, ZHANG M, TAKAGI T. Identity-based partially blind signature in the standard model for electronic cash. Mathematical and Computer Modelling, 2013, 58(1):196-203) and Zhang's scheme (ZHANG J, SUN Z. An ID-based server-aided verification short signature scheme avoid key escrow. Journal of Information Science and Engineering, 2013, 29(3):459-473).

A modified Artificial Bee Colony (ABC) algorithm was proposed for numerical function optimization in this paper, in order to solve the problems of slow convergence and low computational precision of conventional ABC algorithm. The modified ABC algorithm can adjust the step size of the selected neighbor food source position adaptively according to the objective function. On the other hand, the searching method based on a nonlinear adjustment of search range depending on the iteration was introduced for scout bees. The modified ABC algorithm can improve the exploitation, and avoids the premature convergence effectively. The experimental results on six benchmark functions show that, the modified ABC algorithm significantly improves the optimization ability. The modified ABC algorithm can achieve the global minimum values for numerous multimodal functions with high dimension. Compared to the other approaches, the proposed method not only obtains higher quality solutions, but also has a faster convergence speed.

A new broadcast protocol named as location aided broadcast protocol(LABP),which was aided by location information,was put forward.In this protocol,a relay node separated adjacent area into grids,found the broadcast relay gateways(BRGs) with aid of the grids,and broadcasted the message to all its neighbors,with the information of the BRGs.The simulation results show the excellent broadcast performance with LABP.

A time synchronization protocol based on hierarchical structure was designed. In this protocol, at first a hierarchical structure tree was established with sink node as root node, then a pair wise synchronization was performed along the tree to establish a global timescale throughout the sensor network. The simulation results show that the time synchronization protocol can be used for wireless sensor network applications.

Allowing for the shortcomings of existing network security management, a network security management platform based on XMLRPC was designed. The problems of the standardization and combination of log and data in network security management platform were analyzed, a solution method was put forward and the process of implementation was discussed in details. This platform can manage all kinds of network security products running on different operating systems. Moreover, all kinds of log data can be analyzed identically by this platform.

Most of diffusion layers are linear transformations on the vector space GF(2 m) n for SPN structures, which correspond to n-rank matrices under certain bases. The diffusion layers in which branch numbers B equals n+1 are optimal, iff their corresponding matrices have no singular square submatrices. An algorithm was proposed to construct optimal linear layers. In order to validate the optimization of diffusion layers, an algorithm was provided. As an example, a optimal linear mapping over GF(2 8) 8 and its optimization-validation were presented.

In many cases, mobile host needs to obtain accurate or approximate location information in the network, i.e. needs a location service. Five existing location services known as Location Information Flooding(LIF), DREAM Location Service(DLS), Simple Location Service(SLS), Reactive Location Service(RLS), and GLS Location Service were briefly discussed and their performances were compared. Based on these protocols, a new location service named Simple New Location Update Service (SNLUS) was put forward. It improved the Simple Location Service (SLS) by adding "sign of information update" in location information table. Simulation results show that it provides an effective location information service with less overhead.